Basin ecological environment evolution analysis method and system combined with remote sensing image

Abstract

The invention provides a drainage basin ecological environment evolution analysis method and system combined with a remote sensing image, and relates to the technical field of remote sensing. Firstly, a multi-temporal remote sensing image data set covering a target drainage basin and provided with time stamps is obtained, and then ecological feature extraction is carried out on the multi-temporal remote sensing image data set; spatial distribution features and time sequence features of drainage basin ecological elements are obtained, then ecological evolution mode analysis is executed based on the spatial distribution features and the time sequence features, and ecological state transition features of different time nodes are generated; the method comprises the following steps of: determining a drainage basin ecological environment evolution trend including stability, degeneration and restorability directions according to an ecological state transition characteristic, and finally generating a drainage basin ecological environment evolution analysis result containing a time node corresponding relation, so as to obtain a drainage basin ecological environment evolution analysis result. The method is used for indicating a historical evolution process and current state positioning information of a watershed ecological environment.

Description

Technical Field

[0001] The present invention relates to the field of remote sensing technology, and in particular to a method and system for analyzing the evolution of a watershed ecological environment in combination with remote sensing images. Background Art

[0002] In the field of watershed ecological and environmental management and research, accurately understanding the evolution of a watershed's ecological environment is crucial for formulating scientifically sound conservation and development strategies. Traditionally, watershed ecological and environmental analysis has relied primarily on field surveys and fixed-point monitoring data. While field surveys can obtain relatively detailed and accurate firsthand information, they are often limited by human, material, and time resources, making it difficult to cover the entire watershed. Especially for large watersheds, surveys are costly and inefficient. Fixed-point monitoring can only reflect ecological and environmental changes in specific areas and cannot fully capture the dynamic evolution of the entire watershed. Furthermore, data obtained by traditional methods are often not continuous over time, making it difficult to capture subtle trends and key turning points in ecological and environmental changes. With the continuous advancement of remote sensing technology, although some studies have attempted to use remote sensing imagery for watershed ecological and environmental analysis, most have limited their focus on analyzing a single period or simple indicators. They lack the comprehensive utilization of multi-temporal remote sensing imagery and the in-depth exploration of ecological and environmental evolution patterns, failing to accurately reveal the evolving trends and underlying laws of a watershed's ecological environment. Summary of the Invention

[0003] In view of the above-mentioned problems, in combination with the first aspect of the present invention, an embodiment of the present invention provides a method for analyzing the evolution of a watershed ecological environment using remote sensing images, the method comprising: Acquire a multi-temporal remote sensing image data set covering a target watershed, wherein the multi-temporal remote sensing image data set comprises a plurality of remote sensing image units that are continuously acquired and have time stamps; Performing ecological feature extraction processing on the multi-temporal remote sensing image data set to obtain spatial distribution characteristics and time series characteristics of watershed ecological elements in the remote sensing image units, wherein the ecological elements include surface cover type, vegetation cover status, and water body distribution range; Performing ecological evolution pattern analysis based on the spatial distribution characteristics and time series characteristics to generate ecological state transition characteristics of the target watershed at different time nodes, wherein the ecological state transition characteristics include the surface cover type conversion path, the vegetation cover status change gradient, and the water body distribution range expansion and contraction trajectory; Determining the ecological environment evolution trend of the target watershed according to the ecological state transition characteristics, wherein the ecological environment evolution trend includes a stability evolution direction, a degradation evolution direction, and a restoration evolution direction; Based on the ecological environment evolution trend, a basin ecological environment evolution analysis result containing a time node correspondence is generated. The basin ecological environment evolution analysis result is used to indicate the historical evolution process and current state positioning information of the target basin ecological environment.

[0004] On the other hand, an embodiment of the present invention also provides a watershed ecological environment evolution analysis system combined with remote sensing images, including a processor and a machine-readable storage medium, the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the above method.

[0005] Based on the above aspects, by obtaining a multi-temporal remote sensing image data set covering the target watershed and performing ecological feature extraction and processing on the multi-temporal remote sensing image data, the spatial distribution characteristics and time series characteristics of the watershed ecological elements (surface cover type, vegetation coverage status and water body distribution range) can be accurately obtained, reflecting the dynamic changes of the watershed ecological environment from multiple dimensions. Based on the spatial distribution characteristics and time series characteristics, ecological evolution pattern analysis and processing are performed to generate the ecological state transfer characteristics of the target watershed at different time nodes, deeply revealing the inherent laws and key paths of the watershed ecological environment changes, and determining the ecological environment evolution trend according to the ecological state transfer characteristics, including the stability, degradation and recovery evolution directions. Finally, the watershed ecological environment evolution analysis results containing the corresponding relationship between time nodes are generated, which can clearly indicate the historical evolution process of the watershed ecological environment and the current state positioning information. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 It is a schematic diagram of the execution flow of the method for analyzing the evolution of the watershed ecological environment combined with remote sensing images provided in an embodiment of the present invention.

[0007] Figure 2 It is a schematic diagram of exemplary hardware and software components of a watershed ecological environment evolution analysis system combined with remote sensing images provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0008] The present invention will be described in detail below with reference to the accompanying drawings. Figure 1 This is a flow chart of a method for analyzing the evolution of a watershed ecological environment in combination with remote sensing images provided by an embodiment of the present invention. The method for analyzing the evolution of a watershed ecological environment in combination with remote sensing images is introduced in detail below.

[0009] Step S110: Acquire a multi-temporal remote sensing image data set covering the target watershed, wherein the multi-temporal remote sensing image data set includes a plurality of remote sensing image units with time stamps that are continuously acquired.

[0010] In the scenario of this embodiment, multi-temporal remote sensing image data sets are typically acquired through satellite remote sensing platforms or aerial remote sensing equipment. Satellite remote sensing platforms offer wide coverage and periodic, repeated observations, enabling continuous monitoring of target watersheds at set intervals. Aerial remote sensing equipment can capture high-resolution images of target watersheds at specific times.

[0011] During the data collection process, each remote sensing image unit is assigned a time stamp. This time stamp serves as an important basis for subsequent analysis of the temporal evolution of the target watershed's ecological environment. For example, over a long period of time, imagery is collected at fixed intervals, with each acquired image bearing time information such as year, month, day, hour, and minute. These multiple time-stamped remote sensing image units collectively constitute a multi-temporal remote sensing image dataset. This multi-temporal remote sensing image dataset allows for the acquisition of imagery information about the target watershed at different points in time.

[0012] Step S120: performing ecological feature extraction processing on the multi-temporal remote sensing image data set to obtain spatial distribution characteristics and time series characteristics of watershed ecological elements in the remote sensing image unit, wherein the ecological elements include surface cover type, vegetation coverage status and water body distribution range.

[0013] After obtaining the multi-temporal remote sensing image data set, the next step is to extract the ecological features. This is because the remote sensing images contain rich information about the ecological elements of the target watershed, but this information is implicit in the image data and needs to be extracted. Ecological elements mainly include surface cover type, vegetation coverage and water body distribution range, and the features extracted in this embodiment are divided into spatial distribution features and time series features. Spatial distribution features reflect the spatial position and distribution of ecological elements in the target watershed, while time series features reflect the changes of these ecological elements over time.

[0014] Step S121: performing band combination processing on the multi-temporal remote sensing image data set to generate a multispectral image combination unit including a visible light band, a near infrared band, and a short-wave infrared band.

[0015] When extracting ecological features, the first step is to perform band combination processing on multi-temporal remote sensing image data sets. Different bands have different response characteristics to different ground objects. By combining visible light bands, near-infrared bands, and short-wave infrared bands, more comprehensive ground object information in the target watershed can be obtained. The visible light band provides color and texture information of ground objects, which helps to identify different types of surface cover. The near-infrared band has a high reflectivity for vegetation and can well reflect vegetation growth. The short-wave infrared band is more sensitive to water and soil moisture content and can be used to identify water distribution and soil moisture.

[0016] During the band combination process, the individual band data for each remote sensing image unit must be screened and integrated. First, the visible, near-infrared, and short-wave infrared band data for each image unit must be accurate and complete. These band data are then combined to generate a multispectral image combination unit.

[0017] Step S122: performing ground object classification processing on the multispectral image combination unit to identify the boundary contour information of the surface cover type, vegetation coverage area and water distribution area in the remote sensing image unit.

[0018] After obtaining the multispectral image unit, it is necessary to classify the objects in the remote sensing image unit to identify the surface cover type, vegetation coverage area, and boundary contour information of the water distribution area. Object classification is a complex process that requires comprehensive consideration of multiple factors.

[0019] Step S1221: noise suppression is performed on the multispectral image combination unit, and pixel-level classification is performed on the noise-suppressed multispectral image combination unit using a maximum likelihood classification algorithm to generate a classification result map including surface cover type labels, vegetation cover labels, and water body cover labels.

[0020] Before object classification, noise suppression must be performed on the multispectral image combination unit. Remote sensing images may be subject to interference from various factors during acquisition and transmission, resulting in the presence of noise. This noise can affect the accuracy of object classification, so a noise suppression algorithm is needed to mitigate its effects. Common noise suppression algorithms include mean filtering and median filtering.

[0021] After noise suppression processing, the maximum likelihood classification algorithm is used to perform pixel-level classification processing on the image. The maximum likelihood classification algorithm is a classification method based on statistical probability, which assumes that the reflectivity of each ground object category in each band obeys a certain probability distribution. By calculating the probability that each pixel belongs to a different ground object category, the pixel is divided into the category with the highest probability. In this embodiment, the ground object categories mainly include surface cover type, vegetation cover and water body cover. For each pixel, based on its reflectivity value in the visible light band, near infrared band and shortwave infrared band, the probability of it belonging to the ground cover type, vegetation cover area or water body distribution area is calculated, and then it is assigned a corresponding label. Finally, a classification result map containing surface cover type labels, vegetation cover labels and water body cover labels is generated.

[0022] Step S1222: Perform morphological closing operation on the classification result image, connect adjacent pixels of the same label, eliminate small holes and isolated pixels in the classification result, extract edge pixel points of different label areas in the classification result image, and generate boundary contour coordinate sequences of the surface cover type, vegetation coverage area and water distribution area through edge tracking algorithm.

[0023] After generating the classification result map, a morphological closing operation is performed on it to make the classification results more accurate and continuous. Morphological closing is a morphologically based image processing method that connects adjacent pixels of the same label, eliminating small holes and isolated pixels in the classification results. This closing operation can make the boundaries between land cover types, vegetation cover areas, and water distribution areas smoother and more continuous.

[0024] Next, we extract the edge pixels of the different label regions in the classification result image. Edge pixels are pixels located on the boundary between different feature categories and have different characteristics from adjacent pixels. By detecting these edge pixels, we can determine the boundary between different feature categories.

[0025] Finally, an edge tracing algorithm is used to process the edge pixels, generating a coordinate sequence of boundary outlines for land cover types, vegetation cover areas, and water distribution areas. The edge tracing algorithm traverses the edge pixels in a predefined order, connecting them into a continuous boundary outline and recording the coordinate information of each boundary point. This coordinate sequence accurately describes the boundary location of each feature category.

[0026] Step S1223: performing smoothing filtering on the boundary contour coordinate sequence to obtain continuous and smooth boundary contour information.

[0027] After obtaining the boundary contour coordinate sequence, smoothing filtering is required because edge tracing algorithms may produce some local fluctuations and discontinuities. The goal of smoothing filtering is to eliminate noise and irregular fluctuations in the boundary contour, making it more continuous and smooth. Common smoothing filtering algorithms include Gaussian filtering and spline interpolation.

[0028] By smoothing and filtering, the boundary contour coordinate sequence is optimized to obtain continuous and smooth boundary contour information, which can more accurately reflect the actual boundaries of surface cover types, vegetation coverage areas, and water distribution areas.

[0029] Step S123: Calculate the area ratio parameter of the surface cover type, the continuous distribution length parameter of the vegetation coverage area, and the shape complexity parameter of the water body distribution area based on the boundary contour information as the spatial distribution characteristics of the ecological elements.

[0030] Once the continuous and smooth boundary contours are obtained, the spatial distribution characteristics of ecological elements can be calculated based on this information. Spatial distribution characteristics mainly include the area ratio parameters of surface cover types, the continuous distribution length parameters of vegetation cover areas, and the shape complexity parameters of water body distribution areas.

[0031] Step S1231: Count the number of pixels enclosed by the boundary contour information corresponding to the surface cover type, and calculate the actual area value of the surface cover type in combination with the spatial resolution parameter of the remote sensing image unit.

[0032] To calculate the area percentage parameter for a land cover type, we first need to count the number of pixels enclosed by the boundary outline information for the land cover type. Each pixel in a remote sensing image represents a specific area, the size of which is determined by the spatial resolution parameter of the remote sensing image unit. The spatial resolution parameter indicates the actual distance on the ground corresponding to each pixel. For example, a pixel may represent an area of ​​1 meter by 1 meter on the ground.

[0033] By counting the number of pixels within the boundary outline and combining it with the spatial resolution parameter, the actual area value of the surface cover type can be calculated. Specifically, the actual area of ​​the surface cover type can be obtained by multiplying the number of pixels by the actual area represented by each pixel.

[0034] Step S1232: Calculate the total area value of the target watershed, perform a ratio operation on the actual area value of the surface cover type and the total area value of the target watershed, and obtain an area ratio parameter of the surface cover type.

[0035] After obtaining the actual area value of the land cover type, it is necessary to calculate the total area value of the target watershed. The total area of ​​the target watershed can be calculated by counting the number of all pixels in the entire remote sensing image unit and combining it with the spatial resolution parameter.

[0036] Then, the actual area of ​​the land cover type is compared with the total area of ​​the target watershed to obtain the area ratio parameter of the land cover type. This area ratio parameter reflects the proportion of the land cover type in the target watershed.

[0037] Step S1233: extracting the continuous line segment portion in the vegetation coverage area boundary contour information, calculating the pixel length of the continuous line segment and converting it into an actual length value as the continuous distribution length parameter of the vegetation coverage area.

[0038] For vegetation-covered areas, it is necessary to extract the continuous line segments in their boundary contour information. Continuous line segments refer to uninterrupted line segments in the boundary contour, which reflect the continuous distribution of vegetation-covered areas.

[0039] First, continuous line segments are identified from the boundary contours of the vegetation-covered area. Then, the pixel lengths of these continuous line segments are calculated. Similarly, the pixel lengths are converted to actual length values ​​based on the spatial resolution of the remote sensing image unit. This actual length value is the continuous distribution length parameter of the vegetation-covered area, which reflects the continuous growth and distribution of vegetation within the target watershed.

[0040] Step S1234: Calculate the ratio of the perimeter value to the area value of the boundary contour information of the water body distribution area, and use the ratio as a shape complexity parameter of the water body distribution area.

[0041] For water distribution areas, it is necessary to calculate its shape complexity parameter. The shape complexity parameter can reflect the regularity of the water body's shape and is obtained by calculating the ratio of the perimeter value to the area value of the water body distribution area boundary contour information.

[0042] First, calculate the perimeter of the water body's boundary outline—that is, the total length of the boundary outline. Then, calculate the area of ​​the water body using a method similar to that used to calculate the area of ​​the land cover type. Finally, perform a ratio operation on the perimeter and area values ​​to obtain the shape complexity parameter for the water body. A larger shape complexity parameter indicates a more complex water body shape, potentially with more bends and branches. A smaller shape complexity parameter indicates a more regular water body shape.

[0043] Step S1235: performing standardized splicing processing on the area ratio parameter, the continuous distribution length parameter and the shape complexity parameter to generate a spatial distribution feature set.

[0044] After calculating the area ratio parameters of the surface cover type, the continuous distribution length parameters of the vegetation coverage area, and the shape complexity parameters of the water distribution area, these parameters need to be standardized and spliced ​​to generate a set of spatial distribution features.

[0045] Since these parameters may have different dimensions and ranges, they need to be normalized to facilitate subsequent analysis and comparison. Normalization can convert each parameter to the same range and eliminate the influence of dimensions.

[0046] Then, the standardized parameters are spliced ​​together to generate a spatial distribution feature set, which includes spatial distribution feature information of land cover types, vegetation coverage areas, and water body distribution areas.

[0047] Step S124: extracting multispectral reflectance variation information of the same location point in remote sensing image units with different time tags, and generating a time series reflectance curve of the surface cover type, vegetation coverage status and water body distribution range.

[0048] In addition to spatial distribution characteristics, it is also necessary to extract the time series characteristics of ecological elements. Time series characteristics can reflect the changes of ecological elements over time. By extracting the multispectral reflectance variation information of remote sensing image units at different time marks at the same location, a time series reflectance curve is generated.

[0049] Step S1241: selecting a plurality of evenly distributed sampling points in the multi-temporal remote sensing image data set, wherein the sampling points cover the land cover type area, vegetation cover area and water body distribution area.

[0050] To comprehensively capture the temporal characteristics of ecological elements, it is necessary to select multiple, evenly distributed sampling points from a multi-temporal remote sensing image dataset. These sampling points should cover areas of land cover type, vegetation cover, and water distribution to ensure that reflectance variation information for different ecological elements can be captured.

[0051] When selecting sampling points, consider the topography, landforms, and distribution of land features within the target watershed, striving to ensure that sampling points are evenly distributed throughout the entire basin. A grid sampling approach can be employed, dividing the target watershed into several small grids and selecting a sampling point within each grid. This ensures a representative distribution of sampling points and accurately reflects changes in ecological factors within the target watershed.

[0052] Step S1242: for each sampling point, extracting the visible light band reflectance value, near infrared band reflectance value, and shortwave infrared band reflectance value of the corresponding pixel in the remote sensing image unit marked at different times.

[0053] After selecting sampling points, for each sampling point, we need to extract the reflectance values ​​of the visible light band, near-infrared band, and short-wave infrared band for the corresponding pixels in the remote sensing image units marked at different times. The reflectance values ​​of each sampling point at different times reflect the changes in ecological factors at that location over time.

[0054] When extracting reflectance values, it is important to ensure that the band data for each image unit is accurate and that the corresponding pixel location of each sampling point is correct. The image's geographic coordinates and the sampling point's location information can be used to determine the corresponding pixel location of each sampling point in each image unit, and then the reflectance value of the corresponding band can be extracted from that pixel.

[0055] Step S1243: Arrange the visible light band reflectance values ​​of the same sampling point at different time marks in chronological order to generate a visible light band time series reflectance curve.

[0056] By arranging the visible light reflectance values ​​of the same sampling point at different time stamps in chronological order, we can generate a visible light reflectance time series curve for that sampling point. This curve reflects the temporal changes in the surface cover type, vegetation cover, and water distribution range of that sampling point in the visible light band.

[0057] When arranging reflectance values, they must be arranged in strict chronological order to ensure that the curve accurately reflects the temporal evolution of ecological factors. The horizontal axis of the curve represents time, and the vertical axis represents the reflectance value in the visible light band. By observing the shape and trend of the curve, we can analyze the changes in ecological factors in the visible light band.

[0058] Step S1244: Arrange the near-infrared band reflectance values ​​of the same sampling point at different time marks in chronological order to generate a near-infrared band time series reflectance curve.

[0059] Similarly, the near-infrared reflectance values ​​at the same sampling point at different time stamps are arranged in chronological order to generate a near-infrared time series reflectance curve. The near-infrared band is sensitive to vegetation growth, so this curve can well reflect the changes in vegetation cover at the sampling point over time.

[0060] By observing the changing trends of the near-infrared band time series reflectance curve, we can understand the growth cycle, health status, and changes in the coverage of vegetation. For example, if the curve shows an upward trend, it means that the vegetation is growing well and the coverage may be expanding; if the curve shows a downward trend, it may indicate that the vegetation has been disturbed or damaged.

[0061] Step S1245: Arrange the shortwave infrared band reflectance values ​​of the same sampling point at different time marks in chronological order to generate a shortwave infrared band time series reflectance curve.

[0062] The shortwave infrared reflectance values ​​at the same sampling point at different timestamps are arranged in chronological order to generate a shortwave infrared time series reflectance curve. The shortwave infrared band is sensitive to water and soil moisture content, and this curve can reflect the distribution of water bodies and changes in soil moisture over time at the sampling point.

[0063] By observing changes in shortwave infrared time series reflectance curves, we can analyze the fluctuations in water levels and the changing trends in soil moisture. For example, a sudden increase in the reflectance value of the curve may indicate precipitation or water inflow in the area; a continuous decrease in the reflectance value may indicate a decrease in water levels or evaporation of soil moisture.

[0064] Step S1246: combining the visible light band time series reflectance curve, the near infrared band time series reflectance curve, and the shortwave infrared band time series reflectance curve to generate a multi-band time series reflectance curve set of the surface cover type, vegetation coverage status, and water body distribution range.

[0065] Finally, the time series reflectance curves for the visible light band, the near-infrared band, and the shortwave infrared band are combined to generate a multi-band time series reflectance curve set that reflects land cover type, vegetation cover, and water distribution. This multi-band time series reflectance curve set contains more information and can more comprehensively reflect the changes in ecological factors over time.

[0066] By analyzing a collection of multi-band time series reflectance curves, we can comprehensively consider the information of different bands and more accurately judge the transition of surface cover types, changes in vegetation cover conditions, and the expansion or contraction of water body distribution range.

[0067] Step S130: Perform ecological evolution pattern analysis and processing based on the spatial distribution characteristics and time series characteristics to generate ecological state transition characteristics of the target watershed at different time nodes. The ecological state transition characteristics include the surface cover type conversion path, the vegetation cover condition change gradient, and the water body distribution range expansion and contraction trajectory.

[0068] After extracting the spatial distribution and time series characteristics of watershed ecological elements, in order to gain a deeper understanding of the dynamic changes in the target watershed's ecological environment over time, it is necessary to conduct ecological evolution pattern analysis and processing to obtain the ecological state transition characteristics at different time points. These ecological state transition characteristics can intuitively reflect the evolutionary laws of the ecosystem.

[0069] Step S131: performing time dimension alignment processing on the spatial distribution features, arranging the spatial distribution features of different time tags in chronological order, and generating a spatial feature time series.

[0070] Spatial distribution features are collected at different times, and their order may be disordered due to factors such as the collection process and data transmission. To effectively analyze the changes of ecological factors over time, time dimension alignment is required.

[0071] First, a unified standard for time stamping is needed. Different remote sensing image acquisition devices or data processing systems may use different time stamp formats, such as timestamps and date strings. Converting these different time stamp formats into a unified, comparable format, such as converting date strings to timestamps, facilitates subsequent sorting operations.

[0072] Then, a storage structure, such as a list or array, is constructed to store the spatial distribution features and their corresponding time stamps. All spatial distribution feature data are traversed, and each feature and its time stamp are stored as an element in the storage structure.

[0073] Next, use a sorting algorithm to sort the stored structure. Efficient sorting algorithms such as quick sort or merge sort can be used to sort the elements in the order of their time stamps. During the sorting process, ensure that each spatial distribution feature remains associated with its corresponding time stamp to avoid information loss or errors.

[0074] After sorting, a spatial feature time series is generated. Each element in this spatial feature time series corresponds to a specific time node and corresponding spatial distribution characteristics, arranged in chronological order. By observing this spatial feature time series, we can effectively detect the changing trends of spatial distribution characteristics over time, such as the increase or decrease in the proportion of surface cover type area, the change in the continuous distribution length of vegetation cover area, and the changes in the shape complexity of water distribution area.

[0075] Step S132: performing spatial dimension alignment processing on the time series features, arranging the time series features of different sampling points according to spatial positions, and generating a time feature spatial sequence.

[0076] The time series characteristics come from sampling points at different spatial locations. In order to analyze the changing patterns of ecological factors in the spatial dimension, spatial dimension alignment is required.

[0077] The first step is to accurately obtain the spatial location information of each sampling point. This information can be obtained using Geographic Information Systems (GIS) technology and is typically expressed as geographic coordinates (such as longitude and latitude). When obtaining spatial location information, its accuracy and precision must be ensured to avoid inaccuracies in subsequent analysis due to errors in location information.

[0078] The second step is to design a spatial sorting rule based on the spatial location information of the sampling points. This can be done by determining a reference point and direction based on the geographic scope of the target watershed. For example, using the upper left corner of the watershed as the reference point, the sampling points can be sorted from left to right and from top to bottom. Alternatively, the target watershed can be divided into several regions based on a geographic grid, and the regions can be sorted first, followed by sorting the sampling points within each region.

[0079] The third step is to rearrange the time series features of different sampling points according to the designed spatial sorting rules. The time series features of each sampling point are associated with its corresponding spatial location information to ensure that the correspondence between features and locations remains unchanged during the sorting process.

[0080] Finally, a spatial sequence of temporal features is generated. Each element in this spatial sequence represents the temporal characteristics of a sampling point, arranged in order of spatial location. By analyzing this spatial sequence, we can reveal the spatial correlation and heterogeneity of ecological elements. For example, similar temporal series characteristics at adjacent sampling points may indicate that the ecological environment in these areas has similar patterns of change; however, significant differences in characteristics between adjacent sampling points may indicate the presence of different ecological influencing factors, such as topography, soil conditions, or human disturbance.

[0081] Step S133: Calculate the spatial distribution feature difference values ​​of adjacent time nodes in the spatial feature time series to identify the change direction and change amplitude of the surface cover type, vegetation coverage area and water body distribution area.

[0082] Step S1331: extracting the surface cover type area ratio parameter, the vegetation cover area continuous distribution length parameter and the water body distribution area shape complexity parameter of the previous time node in the spatial feature time series.

[0083] In a spatial feature time series, each time node corresponds to a complete set of spatial distribution features. To calculate the difference between adjacent time nodes, it is necessary to accurately extract the relevant parameters of the previous time node.

[0084] For the land cover type area percentage parameter, find the land cover type area percentage data corresponding to the previous time node in the time series. This data was calculated during the previous ecological feature extraction process and stored in the corresponding element of the time series. During extraction, ensure the accuracy of the data to avoid mis-extraction or data corruption.

[0085] For the continuous distribution length parameter of vegetation coverage, the parameter value at the previous time node is also located in the time series. This parameter reflects the continuous growth and distribution of vegetation at that time node.

[0086] For the water body distribution area shape complexity parameter, the value corresponding to the previous time node is extracted from the time series. This water body distribution area shape complexity parameter can reflect the degree of shape regularity of the water body. By analyzing its changes, we can understand the changes in the water body's ecological environment.

[0087] Step S1332: extracting the surface cover type area ratio parameter, the vegetation cover area continuous distribution length parameter and the water body distribution area shape complexity parameter at the next time node in the spatial feature time series.

[0088] Similar to extracting parameters for the previous time node, find the parameters for the land cover type area ratio, the continuous distribution length of vegetation cover area, and the shape complexity of water body distribution area corresponding to the next time node in the spatial feature time series. During the extraction process, ensure the integrity and accuracy of the data and keep it consistent with the parameter extraction method for the previous time node.

[0089] Step S1333: Calculate the difference between the area ratio parameters of the surface cover type at the next time node and the previous time node, determine the expansion direction or contraction direction of the surface cover type according to the positive or negative sign of the difference, and determine the change amplitude of the surface cover type according to the absolute value of the difference.

[0090] Subtract the land cover type area share parameter at the previous time point from the next time point to obtain the difference. A positive difference indicates that the land cover type area share increased between the two time points, indicating an expansion trend. A negative difference indicates that the land cover type area share decreased, indicating a contraction trend.

[0091] The absolute value of the difference reflects the magnitude of the change in land cover type. Larger absolute values ​​indicate a more pronounced change in the area share of a land cover type. For example, a large absolute value of the difference may indicate a large-scale land use change within a short period of time, such as deforestation for agricultural development or urban construction.

[0092] Step S1334: Calculate the difference between the continuous distribution length parameters of the vegetation coverage area at the next time node and the previous time node, determine the direction of connectivity enhancement or weakening of the vegetation coverage area according to the positive or negative sign of the difference, and determine the change amplitude of the vegetation coverage area according to the absolute value of the difference.

[0093] Calculate the difference between the continuous length parameter of the vegetation-covered area at the previous and next time points. A positive difference indicates that the continuous length of the vegetation-covered area increased between the two time points, indicating that the connectivity of the vegetation-covered area has increased. This may be due to factors such as natural vegetation growth, ecological restoration projects, or favorable climatic conditions.

[0094] If the difference is negative, it means that the continuous distribution length of the vegetation cover area has decreased and the connectivity has weakened. This may be because human activities such as deforestation and road construction have fragmented the vegetation into small patches, reducing the connectivity of the vegetation.

[0095] The absolute value of the difference reflects the magnitude of the change in connectivity within the vegetation cover area. Larger absolute values ​​indicate more significant changes in vegetation connectivity and potentially greater impacts on the ecosystem.

[0096] Step S1335: Calculate the difference in the shape complexity parameters of the water distribution area between the next time node and the previous time node, determine the shape regularization direction or irregularization direction of the water distribution area according to the positive or negative sign of the difference, and determine the change amplitude of the water distribution area according to the absolute value of the difference.

[0097] Calculate the difference in the shape complexity parameter of the water distribution area between the previous and next time points. A positive difference indicates an increase in the shape complexity of the water distribution area. This means the shape of the water distribution area has become more complex, possibly with more bends and branches. This could be due to river diversion, water conservancy project construction, or soil erosion.

[0098] If the difference is negative, it means that the shape complexity of the water body distribution area has decreased and the shape has become more regular. This may be due to water body management projects, such as river channel regulation and wetland restoration, which have made the shape of the water body more regular.

[0099] The absolute value of the difference reflects the degree of change in the shape of the water body's distribution area. The larger the absolute value, the more significant the shape change, and the greater the impact on the aquatic ecosystem, such as the habitat and breeding environment of aquatic organisms.

[0100] Step S134: Calculate the time series feature difference values ​​of adjacent spatial positions in the time feature space sequence to identify the heterogeneous distribution pattern of the ecological elements in the spatial dimension.

[0101] In a temporal feature spatial sequence, the time series features corresponding to sampling points at adjacent spatial locations have a certain correlation. By calculating the difference values ​​between these adjacent features, we can discover the spatial variation trends and heterogeneity of ecological factors.

[0102] First, determine the definition of adjacent sampling points. Depending on the geographic scope of the target watershed and the density of sampling points, different adjacency rules can be defined. For example, on a two-dimensional plane, the closest sampling points can be defined as adjacent, or based on a geographic grid, sampling points within the same grid can be considered adjacent.

[0103] Then, for each sampling point's time series characteristics, compare them with those of adjacent sampling points. This comparison can include differences in multispectral reflectance curves, differences in the timing of ecological changes, and so on. For example, using multispectral reflectance curves, one can calculate the reflectance difference between two curves at the same wavelength and time point, or calculate an overall similarity metric, such as a correlation coefficient.

[0104] Next, statistical analysis is performed on the calculated differences. Statistics such as the mean, standard deviation, maximum, and minimum values ​​can be calculated to understand the distribution of these differences. By analyzing these statistics, patterns in the spatial heterogeneity of ecological factors can be revealed. For example, a large standard deviation indicates significant differences in characteristics between adjacent sampling points, indicating strong spatial heterogeneity of ecological factors; conversely, a low standard deviation indicates weak heterogeneity.

[0105] Finally, based on the distribution of difference values, we can summarize the spatial heterogeneity of ecological factors. We can create a spatial distribution map of difference values ​​to visually demonstrate the degree of heterogeneity of ecological factors across different regions. For example, we can use different colors or symbols to represent the magnitude of difference values, effectively identifying areas with significant ecological changes and areas with relatively stable ecological factors.

[0106] Step S135: constructing an ecological state transition matrix according to the change direction, change amplitude and heterogeneity distribution law, wherein the ecological state transition matrix is ​​used to represent the ecological state transition probability of the target watershed from one time node to the next time node.

[0107] Step S1351: Define the ecological state set of the target watershed, where the ecological state set includes a stable surface cover state, an expanding surface cover state, a shrinking surface cover state, a stable vegetation connectivity state, an enhanced vegetation connectivity state, a weakened vegetation connectivity state, a regular water body shape state, and an irregular water body shape state.

[0108] To construct the ecological state transition matrix, we first need to identify the possible ecological states of the target watershed. Based on the previously calculated information on the direction and magnitude of change in land cover type, vegetation cover area, and water distribution, we define a series of ecological states.

[0109] A stable land cover state indicates that the area share of a land cover type has not changed significantly over a set period of time, meaning that the difference in area share between adjacent time points is within a small threshold. An expanding land cover state indicates that the area share has increased, exceeding the set expansion threshold; a contracting land cover state indicates that the area share has decreased, exceeding the set contraction threshold.

[0110] The stable state of vegetation connectivity indicates that the continuous distribution length of the vegetation coverage area is relatively stable, and the length difference between adjacent time nodes is within a small threshold range. The increasing state of vegetation connectivity indicates that the length has increased, exceeding the set increasing threshold; the decreasing state of vegetation connectivity indicates that the length has decreased, exceeding the set decreasing threshold.

[0111] The regular shape state indicates that the shape complexity of the water body distribution area is low and relatively stable, and the difference in shape complexity between adjacent time nodes is within a small threshold. The irregular shape state indicates that the shape complexity is high and changing, exceeding the set irregularity threshold. These ecological states cover the main changes in the ecological environment of the target watershed.

[0112] Step S1352: Count the occurrence frequencies of the change direction and change amplitude at historical time nodes, and calculate the number of transitions from each initial ecological state to each target ecological state.

[0113] Conduct a detailed statistical analysis of the direction and magnitude of change at historical time points. For each initial ecological state, count the number of times it transitions to each target ecological state at different time points.

[0114] First, we traverse the spatial feature time series and the temporal feature spatial series and determine the ecological state of each time node based on the defined ecological state set. For each time node, we determine the ecological state of its land cover type, vegetation connectivity, and water body shape.

[0115] Then, for each initial ecological state, record its transitions to each target ecological state at subsequent time points. A two-dimensional array or matrix can be used to store the transition counts, with rows representing initial ecological states and columns representing target ecological states. Each time a transition occurs from an initial ecological state to a target ecological state, the corresponding array element is incremented by 1.

[0116] During the statistical process, we must ensure the accuracy and completeness of the data and carefully record and analyze the changes in ecological status at each time point. By using a large amount of historical data statistics, we can obtain information on the frequency of transitions between various ecological states.

[0117] Step S1353: Divide the number of transitions by the total number of occurrences of the initial ecological state to obtain a transition probability from the initial ecological state to the target ecological state.

[0118] After obtaining the number of transitions from each initial ecological state to the target ecological state, divide it by the total number of occurrences of the initial ecological state to obtain the corresponding transition probability.

[0119] First, count the total number of occurrences of each initial ecological state. This can be done by traversing the two-dimensional array that records the number of transitions and summing the elements in each row to obtain the total number of occurrences of the initial ecological state corresponding to that row.

[0120] Then, for each initial ecological state, the number of transitions to the target ecological state is divided by the total number of occurrences of the initial ecological state to obtain the transition probability. The calculated transition probabilities are stored in a new two-dimensional array or matrix, where the rows of the matrix represent the initial ecological state and the columns represent the target ecological state.

[0121] The transition probability reflects the likelihood of transitioning between different ecological states. For example, if the transition probability from a stable land cover state to an expanding land cover state is high, it means that under certain conditions, the land cover is more likely to transition from a stable state to an expanding state.

[0122] Step S1354: Arrange the transition probabilities in row priority order to generate a two-dimensional matrix structure with row index being the initial ecological state and column index being the target ecological state.

[0123] The calculated transition probabilities are arranged in row-priority order to form a two-dimensional matrix structure.

[0124] First, determine the size of the matrix. The number of rows in the matrix is ​​equal to the number of states in the ecological state set, and the number of columns is also equal to the number of states in the ecological state set.

[0125] Then, the calculated transition probabilities are filled into the matrix in row-priority order. That is, the transition probabilities in the first row are first filled into the first row of the matrix, and then the transition probabilities in the second row are filled into the second row of the matrix, and so on.

[0126] In this matrix, row indices represent initial ecological states, and column indices represent target ecological states. Each element in the matrix represents the probability of transitioning from the corresponding initial ecological state to the target ecological state. For example, the element in the first row and first column of the matrix represents the probability of transitioning from the first initial ecological state to the first target ecological state. This matrix structure allows for intuitive detection of transition relationships and probabilities between various ecological states, facilitating subsequent analysis and calculations.

[0127] Step S1355: normalizing the two-dimensional matrix structure to obtain the ecological state transition matrix.

[0128] In order to make the ecological state transition matrix more comparable and stable, it is necessary to normalize the two-dimensional matrix structure. The purpose of normalization is to ensure that the elements in the matrix meet the set conditions, such as the sum of all row elements is 1.

[0129] First, check the sum of the elements in each row of the matrix. For each row of the matrix, calculate the sum of all the elements in that row.

[0130] Then, for each row of elements, if the sum is not equal to 1, the row of elements is normalized. The normalization method is to divide each element of the row by the sum of the elements of the row so that the sum of the elements of the row is equal to 1 after processing.

[0131] Through standardization, the dimensional differences in the transition probabilities between different ecological states can be eliminated, making the matrix more accurately reflect the ecological state transition of the target watershed. After standardization, the final ecological state transition matrix is ​​obtained.

[0132] Step S136: extracting the surface cover type conversion path, vegetation cover change gradient, and water body distribution range expansion and contraction trajectory based on the ecological state transition matrix as the ecological state transition characteristics.

[0133] The ecological state transition matrix can be used to analyze the transition probability of land cover types between different ecological states. By finding the row and column elements related to land cover in the matrix, the probability of land cover type transitioning from one state to another can be determined.

[0134] First, determine the set of ecological states of land cover types, including forest, farmland, urban construction land, grassland, etc. Then, find the rows and columns corresponding to these land cover states in the ecological state transition matrix.

[0135] For each land cover state, find the elements in that row of the matrix with the highest probability. The land cover states represented by the corresponding columns of these elements are the possible transition targets. For example, if the probability of transitioning from forest to farmland is high in the matrix, then forest to farmland is a possible transition path.

[0136] All possible conversion paths are sorted by their probability to form land cover type conversion paths. These paths can reveal changing trends in land cover types over time, helping to understand the impacts of human activities or natural factors on land cover. For example, an increase in the probability of conversion from forest to urban construction land may indicate that accelerated urbanization is putting pressure on forest resources.

[0137] Based on the elements related to vegetation connectivity in the ecological state transition matrix, the change gradient of vegetation cover between different time nodes was calculated. The change gradient reflects the rate and direction of vegetation cover change.

[0138] First, the set of ecological states related to vegetation connectivity is determined, including the stable vegetation connectivity state, the enhanced vegetation connectivity state, the weakened vegetation connectivity state, etc. Then, the rows and columns corresponding to these states are found in the ecological state transition matrix.

[0139] For adjacent time nodes, the degree of change in vegetation cover is calculated by comparing the transition probabilities of vegetation connectivity states. For example, if the probability of transitioning from an enhanced to a weakened state is high, it indicates that vegetation cover is declining during that time period, with a negative gradient.

[0140] The gradient of change can be quantified by calculating the difference in vegetation connectivity state transition probabilities between different time points. These gradients are arranged in chronological order to form a sequence of vegetation cover change gradients. By analyzing this sequence, we can understand the long-term trends in vegetation cover and the time periods when vegetation is most affected.

[0141] From the elements related to water body shape in the ecological state transition matrix, the expansion and contraction trajectory of water body distribution range can be extracted.

[0142] First, we identify a set of ecological states related to water body shape, including regular and irregular water body shapes, as well as derived states related to the expansion and contraction of water body distribution. Then, we identify the rows and columns corresponding to these states in the ecological state transition matrix.

[0143] For each time point, the elements in the matrix related to the shape and distribution of the water body are analyzed. If the probability of the water body distribution shifting from a contracting state to an expanding state is high, it indicates that the water body is likely to expand after that time point. Conversely, if the probability of the water body shifting from an expanding state to a contracting state is high, it means that the water body is likely to shrink.

[0144] By analyzing multiple consecutive time points, we track changes in the distribution range of water bodies. By connecting the trends in the distribution range of water bodies at each time point, we can form a trajectory of water body expansion and contraction. For example, if, over a series of time points, the matrix shows that the water body is initially in a contraction state, then there is an increase in the probability of transitioning from contraction to expansion, and then a trend of returning to a contraction state, then the expansion and contraction trajectory of the water body distribution range will show a dynamic change of first contraction, then expansion, and finally contraction again.

[0145] During the extraction process, it's important to comprehensively consider the impact of multiple factors on the distribution of water bodies. For example, natural and human factors such as rainfall, river diversion, and water conservancy project construction can alter the distribution of water bodies. The expansion and contraction trajectories can be verified and supplemented with auxiliary information such as historical meteorological data and water conservancy project records, improving their accuracy and reliability.

[0146] Step S140: determining the ecological environment evolution trend of the target watershed according to the ecological state transition characteristics, wherein the ecological environment evolution trend includes a stability evolution direction, a degradation evolution direction, and a restoration evolution direction.

[0147] Step S141: Extract the surface cover type conversion path in the ecological state transition characteristics, analyze the proportion of the surface cover type converted to the first ecological value type and the proportion of the surface cover type converted to the second ecological value type, and the ecological value of the first ecological value type is greater than the ecological value of the second ecological value type.

[0148] After extracting the land cover type conversion paths from the ecological state transition characteristics, these paths need to be analyzed in detail. First, the specific definitions of the first and second ecological value types are clarified. The first ecological value type generally includes land cover types with high ecological service functions, such as forests and wetlands, which can provide important ecological functions such as water conservation, biodiversity protection, and climate regulation. The second ecological value type may be types with relatively low ecological value, such as farmland and urban construction land.

[0149] The extracted land cover type conversion paths were classified and counted. Across all conversion paths, the number of land cover types that converted to the first ecological value type and the number that converted to the second ecological value type were counted. The proportions of these two types of conversions were then calculated. Specifically, the number of conversions to the first ecological value type was divided by the total number of conversion paths to obtain the proportion of conversions to the first ecological value type. Similarly, the proportion of conversions to the second ecological value type was calculated.

[0150] By analyzing the relationship between these two ratios, we can preliminarily determine the direction of impact of changes in land cover types on the ecological environment. If the proportion of conversion to the first ecological value type is high, it indicates that changes in land cover types are conducive to improving the ecological environment. Conversely, if the proportion of conversion to the second ecological value type is high, it may mean that the ecological environment is facing certain pressures.

[0151] Step S142: extracting the vegetation coverage change gradient in the ecological state transition feature, and analyzing the area ratio of regions where the vegetation coverage presents a positive change and the area ratio of regions where the vegetation coverage presents a negative change.

[0152] After extracting the vegetation cover change gradient from the ecological state transition characteristics, it is necessary to further analyze its performance in different areas of the target watershed. First, the target watershed is divided into several small areas, which can be divided according to geographic grids or natural geographic units.

[0153] For each small area, the vegetation cover gradient was used to determine whether the change was positive or negative. Positive changes generally indicate favorable ecological conditions, such as increased vegetation cover and enhanced vegetation connectivity; negative changes indicate unfavorable conditions, such as decreased vegetation cover and decreased vegetation connectivity.

[0154] The number of regions showing positive and negative changes was counted. Based on the area information for each region, the areas showing positive and negative changes were calculated. The proportion of each of these areas to the total area of ​​the target watershed was then calculated.

[0155] By analyzing the relationship between these two area proportions, we can understand the overall trend of vegetation cover within the target watershed. If the area proportion showing positive changes is greater than the area proportion showing negative changes, it indicates that vegetation cover is generally improving; conversely, it indicates that vegetation cover may be deteriorating.

[0156] Step S143: extracting the expansion and contraction trajectory of the water body distribution range in the ecological state transition feature, and analyzing the proportion of the length of the area where the water body distribution range shows an expansion trend and the proportion of the length of the area where the water body distribution range shows a contraction trend.

[0157] After extracting the expansion and contraction trajectories of water distribution ranges from the ecological state transition characteristics, they are analyzed in detail. The water distribution area is divided into several small segments, such as river segments or different areas of lakes.

[0158] For each segment, the expansion and contraction trajectory is used to determine whether its water distribution range is expanding or contracting. The number of segments showing an expansion trend and the number showing a contraction trend are counted. Combined with the length of each segment, the length of the area showing an expansion trend and the length of the area showing a contraction trend are calculated.

[0159] Next, the proportion of each of these two areas' lengths to the total length of the water body is calculated. By comparing these two ratios, we can understand the overall changes in the water body's distribution range within the target basin. If the proportion of the length of the area showing an expansion trend is greater than the proportion of the length of the area showing a contraction trend, it indicates that the water body's distribution range is generally expanding; conversely, it indicates that the water body's distribution range may be shrinking.

[0160] Step S144: Calculate the difference between the conversion ratio of the first ecological value type and the conversion ratio of the second ecological value type as an indicator of the evolution trend of the surface cover ecological environment.

[0161] Subtract the proportion of surface cover type converted to the first ecological value type from the proportion of surface cover type converted to the second ecological value type obtained by calculation above to obtain the surface cover ecological environment evolution trend index.

[0162] If the difference is positive, it indicates that a large proportion of land cover types are shifting toward high-ecological-value types, meaning that changes in land cover are beneficial to the ecological environment and that the land cover ecosystem may be evolving in a positive direction. For example, a large, positive difference may indicate that forest area is increasing and farmland or urban construction land is decreasing, thereby enhancing ecosystem services.

[0163] If the difference is negative, it indicates that a large proportion of land cover types have shifted to low ecological value types, potentially leading to ecological degradation. For example, a negative difference with a large absolute value may indicate that a large amount of forest has been cut down for urban construction or agricultural development, disrupting the ecological balance.

[0164] Step S145: Calculate the difference between the area ratio of the positive change region and the area ratio of the negative change region as the vegetation coverage ecological environment evolution trend indicator.

[0165] The difference between the area proportion of regions with positive changes in vegetation coverage and the area proportion of regions with negative changes was calculated to obtain the vegetation coverage ecological environment evolution trend indicator.

[0166] A positive difference indicates that vegetation cover is generally improving, and the vegetation ecosystem may be recovering. For example, a positive and large difference may indicate good vegetation growth, enhanced vegetation connectivity, and improved ecosystem stability.

[0167] If the difference is negative, it indicates that vegetation cover may be degrading and the vegetation ecosystem is facing certain pressures. For example, a negative difference with a large absolute value may indicate that deforestation, grassland degradation and other problems are more serious, affecting the functioning of the ecosystem.

[0168] Step S146: Calculate the difference between the length ratio of the water body expansion area and the length ratio of the water body contraction area as an indicator of the evolution trend of the water body distribution ecological environment.

[0169] The difference between the length proportion of the area where the water body distribution range shows an expansion trend and the length proportion of the area where the water body distribution range shows a contraction trend is calculated to obtain the water body distribution ecological environment evolution trend index.

[0170] A positive difference indicates that the overall distribution of the water body is expanding, and the ecological environment surrounding the water body may be improving. For example, a positive and large difference may indicate an increase in river flow and an expansion of lake area, which is beneficial for maintaining the living environment of aquatic organisms and the balance of the ecosystem.

[0171] If the difference is negative, it indicates that the distribution range of the water body may be shrinking and the ecological environment of the water body distribution may face certain challenges. For example, a negative difference with a large absolute value may indicate water shortages, river dry-up, and other problems, which may have adverse impacts on the ecological environment and human life.

[0172] Step S147: According to the comprehensive evaluation results of the surface cover ecological environment evolution trend index, the vegetation cover ecological environment evolution trend index and the water body distribution ecological environment evolution trend index, the ecological environment evolution trend of the target watershed is determined to be a stable evolution direction, a degenerative evolution direction or a restorative evolution direction.

[0173] For example, step S1471: set the first threshold range of the surface cover ecological environment evolution trend index, the second threshold range of the vegetation cover ecological environment evolution trend index and the third threshold range of the water body distribution ecological environment evolution trend index. The first threshold range, the second threshold range and the third threshold range all include positive and negative intervals and stable intervals near zero.

[0174] In order to determine the ecological environment evolution trend of the target basin based on the three ecological environment evolution trend indicators, it is necessary to set the threshold range of each indicator.

[0175] For the land cover ecological environment evolution trend indicator, a first threshold range is set based on historical data and the actual ecological environment. This first threshold range includes a positive interval, a negative interval, and a stable interval near zero. A positive interval indicates that the land cover ecological environment is improving, a negative interval indicates that it is deteriorating, and a stable interval indicates that the land cover ecological environment is relatively stable.

[0176] Similarly, the second and third threshold ranges are set for the vegetation cover ecological environment evolution trend index and the water body distribution ecological environment evolution trend index, respectively. The setting of these threshold ranges should fully consider the complexity and uncertainty of the ecosystem to ensure that they can accurately reflect the actual changes in the ecological environment.

[0177] Step S1472: Determine whether the surface cover ecological environment evolution trend index falls within the stable interval of the first threshold range, whether the vegetation cover ecological environment evolution trend index falls within the stable interval of the second threshold range, and whether the water body distribution ecological environment evolution trend index falls within the stable interval of the third threshold range.

[0178] The calculated ecological environment evolution trend indicators of surface cover, vegetation cover and water body distribution are compared with their respective threshold ranges.

[0179] Determine whether the land cover ecological environment evolution trend indicator is within the stable range of the first threshold range. If it is within the stable range, it means that the land cover ecological environment is relatively stable at the current stage, with no obvious signs of improvement or degradation.

[0180] Similarly, it is determined whether the vegetation coverage ecological environment evolution trend index falls within the stable interval of the second threshold range, and whether the water body distribution ecological environment evolution trend index falls within the stable interval of the third threshold range.

[0181] Step S1473: If the three evolution trend indicators all fall within the stable range, it is determined that the ecological environment evolution trend of the target watershed is in a stable evolution direction.

[0182] When all three evolutionary trend indicators fall within their respective stable ranges, the ecological environment of the target basin can be determined to be evolving in a stable direction. This means that at this stage, ecological elements such as land cover, vegetation cover, and water distribution in the target basin are relatively stable, and the ecosystem is in a state of dynamic equilibrium. This stability may be due to the limited impact of human activities or the ecosystem's strong self-regulating ability.

[0183] Step S1474: If at least two evolution trend indicators fall within the negative interval and do not fall within the stable interval, it is determined that the ecological environment evolution trend of the target watershed is in a degenerative evolution direction.

[0184] If at least two trend indicators fall within the negative range of their respective thresholds and outside the stable range, the target basin's ecological environment is showing signs of degradation in multiple aspects. For example, if both the surface cover and vegetation cover trend indicators are negative and outside the stable range, this may indicate a shift in land cover toward a type of low ecological value and a deterioration in vegetation cover, negatively impacting the functioning and stability of the entire ecosystem. In this case, the target basin's ecological environment can be determined to be degrading.

[0185] Step S1475: If at least two evolution trend indicators fall within the positive interval and do not fall within the stable interval, it is determined that the ecological environment evolution trend of the target watershed is a restorative evolution direction.

[0186] When at least two trend indicators fall within the positive range of their respective thresholds and outside the stable range, the target basin's ecological environment is showing signs of recovery and improvement in multiple aspects. For example, if both the vegetation cover and water distribution indicators are positive and outside the stable range, this indicates that vegetation cover is improving and water distribution is expanding, which is conducive to ecosystem recovery and development. In this case, the target basin's ecological environment can be determined to be trending towards recovery.

[0187] Step S1476: If the distribution of each evolution trend indicator does not conform to the above situation, then the dominant direction of the ecological environment evolution trend is comprehensively judged based on the specific numerical values ​​and change rates of the evolution trend indicators.

[0188] If the distribution of each evolution trend indicator does not conform to the above three conditions, that is, none of the three indicators are stable, and at least two indicators are not degraded or recovered, then it is necessary to comprehensively judge the dominant direction of the ecological environment evolution trend based on the specific numerical size and change rate of the evolution trend indicators.

[0189] For example, even if only one indicator is clearly in the positive range, its value is large and its rate of change is rapid, potentially having a significant impact on the overall ecological and environmental evolution. Simultaneously, the values ​​and changes of other indicators, as well as their interrelationships, must be considered. If other indicators, while not in the positive or negative range, also show a slowly changing trend in a certain direction, these factors should also be factored into the overall assessment. Through comprehensive analysis and weighing of various factors, the dominant direction of ecological and environmental evolution can be determined, allowing for the implementation of appropriate ecological protection and management measures.

[0190] Step S150: Generate a basin ecological environment evolution analysis result containing a time node correspondence based on the ecological environment evolution trend, and the basin ecological environment evolution analysis result is used to indicate the historical evolution process and current state positioning information of the target basin ecological environment.

[0191] After determining the ecological environment evolution trend of the target watershed, the next step is to generate the watershed ecological environment evolution analysis results containing the corresponding relationship between time nodes based on this.

[0192] First, it's necessary to integrate the various types of information previously acquired and analyzed. This information includes multi-temporal remote sensing imagery, the spatial distribution characteristics of ecological elements, time series characteristics, ecological state transitions, and ecological and environmental evolution trends. During this integration process, special attention should be paid to the correspondence between each piece of information and time points, as this is the key to constructing a complete evolutionary analysis.

[0193] For ecological state transition characteristics, the specific time points at which each transition occurred must be clearly identified. For example, the time when land cover types transitioned from one to another, the specific manifestations of vegetation cover gradients over time, and the time points corresponding to the expansion and contraction trajectories of water distribution. By mapping these ecological state transition information to time points, the dynamic changes in the ecological environment of the target watershed over time can be effectively demonstrated.

[0194] For ecological and environmental trends, it's important to analyze their phased characteristics within a specific timeframe. If the ecological and environmental trend is one of stability, it's important to examine the time periods during which the ecological environment remained relatively stable, as well as the specific characteristics of each ecological element during that stable state. If the trend is one of degradation, it's important to clarify the onset of degradation, the rate of degradation, and which ecological elements were most impacted. Similarly, for restorative trends, it's important to determine the onset of recovery, the extent of recovery, and which ecological elements played a key role in its recovery.

[0195] When generating analysis results for a watershed's ecological and environmental evolution, a combination of textual description and visual presentation can be used. The textual description should detail the changes in the target watershed's ecological environment at different time points, including specific changes in ecological elements, shifts in ecological status, and the reasons for these trends. For example, describe how, at a specific time point, human activities shifted the land cover from forest to farmland, leading to a decline in vegetation cover and, in turn, affecting the ecological balance of the entire watershed.

[0196] The visualization component can intuitively present the evolution of the ecological environment through methods such as time series graphs and change trajectory graphs. Time series graphs can show the changing trends of certain ecological indicators (such as the proportion of land cover types and the length of continuous distribution of vegetation cover areas) over time, allowing readers to quickly understand the overall direction of ecological evolution. Change trajectory graphs can graphically display the characteristics of ecological state transitions, effectively presenting information such as the transition paths of land cover types, the gradient of vegetation cover changes, and the expansion and contraction trajectories of water distribution.

[0197] Furthermore, to make the analysis more practical, it is also necessary to provide information on the current state of the ecological environment in the target watershed. This can be done by comparing the characteristics of ecological elements at the current time point with historical data to determine the current stage of ecological evolution. For example, if the proportion of forest area in the current land cover type continues to decline, the gradient of vegetation cover change is negative, the distribution range of water bodies shrinks, and the ecological environment evolution trend is in a degenerative direction, then the ecological environment in the target watershed can be judged to be in a degraded stage and appropriate protection and restoration measures are needed.

[0198] Furthermore, when generating analytical results, quality control and verification must be conducted throughout the entire analytical process. The accuracy and reliability of each data source must be checked to ensure the rationality of the calculation process and analytical methods. The credibility of the analytical results can be verified by comparing them with other relevant research or actual monitoring data. If bias or uncertainty is detected in the analytical results, corrections and adjustments must be made promptly to ensure that the results accurately reflect the actual evolution of the ecological environment in the target watershed.

[0199] Figure 2 A schematic diagram illustrates exemplary hardware and software components of a watershed ecological environment evolution analysis system 100 for combining remote sensing imagery with watershed ecological environment evolution analysis, as provided in some embodiments of the present application, that can implement the concepts of the present application. For example, processor 120 can be used in watershed ecological environment evolution analysis system 100 for combining remote sensing imagery with watershed ecological environment evolution analysis, and can be used to perform the functions of the present application.

[0200] The system 100 for analyzing the evolution of the ecological environment of a watershed using remote sensing images can be a general-purpose server or a special-purpose server, both of which can be used to implement the method for analyzing the evolution of the ecological environment of a watershed using remote sensing images described herein. Although only one server is shown in this application, for convenience, the functions described herein can be implemented in a distributed manner on multiple similar platforms to balance the processing load.

[0201] For example, the watershed ecological environment evolution analysis system 100 combined with remote sensing images may include a network port 110 connected to a network, one or more processors 120 for executing program instructions, a communication bus 130, and storage media 140 in different forms, such as a disk, ROM, or RAM, or any combination thereof. Exemplarily, the watershed ecological environment evolution analysis system 100 combined with remote sensing images may also include program instructions stored in ROM, RAM, or other types of non-temporary storage media, or any combination thereof. The method of the present application can be implemented according to these program instructions. The watershed ecological environment evolution analysis system 100 combined with remote sensing images also includes an I / O interface 150 between the computer and other input and output devices.

[0202] For ease of explanation, only one processor is described in the watershed ecological environment evolution analysis system 100 combined with remote sensing images. However, it should be noted that the watershed ecological environment evolution analysis system 100 combined with remote sensing images in this application can also include multiple processors, so the steps performed by one processor described in this application can also be performed jointly or individually by multiple processors. For example, if the processor of the watershed ecological environment evolution analysis system 100 combined with remote sensing images executes step A and step B, it should be understood that step A and step B can also be executed jointly by two different processors or individually in one processor. For example, the first processor executes step A, the second processor executes step B, or the first processor and the second processor execute steps A and B together.

[0203] In addition, an embodiment of the present invention also provides a readable storage medium, which has computer-executable instructions preset therein. When the processor executes the computer-executable instructions, the above-mentioned method for analyzing the evolution of the watershed ecological environment combined with remote sensing images is implemented.

[0204] It should be noted that in order to simplify the description of the present invention and thus help understand one or more embodiments of the invention, in the foregoing description of the embodiments of the present invention, multiple features are sometimes combined into one embodiment, figure or description thereof.

Claims

1. A method for analyzing the evolution of watershed ecological environment using remote sensing images, characterized in that: The method comprises: Acquire a multi-temporal remote sensing image data set covering a target watershed, wherein the multi-temporal remote sensing image data set comprises a plurality of remote sensing image units that are continuously acquired and have time stamps; Performing ecological feature extraction processing on the multi-temporal remote sensing image data set to obtain spatial distribution characteristics and time series characteristics of watershed ecological elements in the remote sensing image units, wherein the ecological elements include surface cover type, vegetation cover status, and water body distribution range; Performing ecological evolution pattern analysis based on the spatial distribution characteristics and time series characteristics to generate ecological state transition characteristics of the target watershed at different time nodes, wherein the ecological state transition characteristics include the surface cover type conversion path, the vegetation cover status change gradient, and the water body distribution range expansion and contraction trajectory; Determining the ecological environment evolution trend of the target watershed according to the ecological state transition characteristics, wherein the ecological environment evolution trend includes a stability evolution direction, a degradation evolution direction, and a restoration evolution direction; Based on the ecological environment evolution trend, a basin ecological environment evolution analysis result containing a time node correspondence is generated. The basin ecological environment evolution analysis result is used to indicate the historical evolution process and current state positioning information of the target basin ecological environment.

2. The method for analyzing the evolution of watershed ecological environment by combining remote sensing images according to claim 1, characterized in that: The ecological feature extraction process is performed on the multi-temporal remote sensing image data set to obtain the spatial distribution characteristics and time series characteristics of the watershed ecological elements in the remote sensing image unit, including: Performing band combination processing on the multi-temporal remote sensing image data set to generate a multispectral image combination unit including a visible light band, a near infrared band, and a short-wave infrared band; Performing ground object classification processing on the multispectral image combination unit to identify the boundary contour information of the surface cover type, vegetation coverage area and water distribution area in the remote sensing image unit; Calculating, based on the boundary contour information, an area ratio parameter of the surface cover type, a continuous distribution length parameter of the vegetation coverage area, and a shape complexity parameter of the water body distribution area as spatial distribution characteristics of the ecological elements; Extracting multispectral reflectance variation information of the same location point in remote sensing image units marked at different times, and generating a time series reflectance curve of the surface cover type, vegetation cover status and water body distribution range; The time series reflectivity curve is subjected to trend analysis and processing, and the rising edge slope parameter, the falling edge slope parameter and the plateau duration parameter of the curve are extracted as the time series characteristics of the ecological element.

3. The method for analyzing the evolution of watershed ecological environment by combining remote sensing images according to claim 2, characterized in that: The performing of ground object classification processing on the multispectral image combination unit to identify the boundary contour information of the surface cover type, vegetation coverage area and water distribution area in the remote sensing image unit includes: Performing noise suppression on the multispectral image combination unit, and performing pixel-level classification on the noise-suppressed multispectral image combination unit using a maximum likelihood classification algorithm to generate a classification result map including surface cover type labels, vegetation cover labels, and water body cover labels; Performing morphological closing operations on the classification result graph to connect adjacent pixels of the same label, eliminating small holes and isolated pixels in the classification result, extracting edge pixel points of different label areas in the classification result graph, and generating boundary contour coordinate sequences of the surface cover type, vegetation cover area, and water distribution area through an edge tracing algorithm; The boundary contour coordinate sequence is smoothed and filtered to obtain continuous and smooth boundary contour information.

4. The method for analyzing the evolution of watershed ecological environment by combining remote sensing images according to claim 2, characterized in that: The calculating, based on the boundary contour information, of the area ratio parameter of the surface cover type, the continuous distribution length parameter of the vegetation coverage area, and the shape complexity parameter of the water body distribution area as the spatial distribution characteristics of the ecological elements includes: Counting the number of pixels enclosed by the boundary contour information corresponding to the surface cover type, and calculating the actual area value of the surface cover type in combination with the spatial resolution parameter of the remote sensing image unit; Calculating the total area value of the target watershed, performing a ratio operation on the actual area value of the surface cover type and the total area value of the target watershed to obtain an area ratio parameter of the surface cover type; Extracting the continuous line segment portion from the boundary contour information of the vegetation covered area, calculating the pixel length of the continuous line segment and converting it into an actual length value as a continuous distribution length parameter of the vegetation covered area; Calculating the ratio of the perimeter value to the area value of the boundary contour information of the water body distribution area, and using the ratio as a shape complexity parameter of the water body distribution area; The area ratio parameter, the continuous distribution length parameter and the shape complexity parameter are standardized and spliced ​​to generate a spatial distribution feature set.

5. The method for analyzing the evolution of watershed ecological environment by combining remote sensing images according to claim 2, characterized in that: The method of extracting multispectral reflectance change information of remote sensing image units at different time marks at the same location point to generate a time series reflectance curve of the surface cover type, vegetation cover status and water body distribution range includes: Selecting a plurality of evenly distributed sampling points in the multi-temporal remote sensing image data set, wherein the sampling points cover the land cover type area, vegetation cover area and water body distribution area; For each sampling point, extract the reflectance value of the visible light band, near infrared band and short wave infrared band of the corresponding pixel in the remote sensing image unit marked at different times; The visible light band reflectance values ​​of the same sampling point at different time marks are arranged in chronological order to generate a visible light band time series reflectance curve; The near-infrared band reflectance values ​​of the same sampling point at different time marks are arranged in chronological order to generate a near-infrared band time series reflectance curve; Arrange the shortwave infrared band reflectance values ​​of the same sampling point at different time marks in chronological order to generate a shortwave infrared band time series reflectance curve; The visible light band time series reflectance curve, the near infrared band time series reflectance curve and the shortwave infrared band time series reflectance curve are combined to generate a multi-band time series reflectance curve set of the surface cover type, vegetation coverage status and water body distribution range.

6. The method for analyzing the evolution of watershed ecological environment by combining remote sensing images according to claim 1, characterized in that: The performing of ecological evolution pattern analysis based on the spatial distribution characteristics and the time series characteristics to generate ecological state transition characteristics of the target watershed at different time nodes includes: Performing time dimension alignment processing on the spatial distribution features, arranging the spatial distribution features of different time tags in chronological order, and generating a spatial feature time series; Performing spatial dimension alignment processing on the time series features, arranging the time series features of different sampling points according to spatial positions, and generating a time feature spatial sequence; Calculating the spatial distribution feature difference values ​​of adjacent time nodes in the spatial feature time series to identify the change direction and magnitude of the surface cover type, vegetation cover area and water body distribution area; Calculating the time series characteristic difference values ​​of adjacent spatial positions in the time characteristic spatial sequence to identify the heterogeneous distribution pattern of the ecological elements in the spatial dimension; Constructing an ecological state transition matrix based on the change direction, change amplitude and heterogeneity distribution law, wherein the ecological state transition matrix is ​​used to represent the ecological state transition probability of the target watershed from one time node to the next time node; Based on the ecological state transition matrix, the surface cover type conversion path, the vegetation cover change gradient and the water body distribution range expansion and contraction trajectory are extracted as the ecological state transition characteristics.

7. The method for analyzing the evolution of watershed ecological environment by combining remote sensing images according to claim 6, characterized in that: The calculating of spatial distribution feature difference values ​​of adjacent time nodes in the spatial feature time series to identify the change direction and magnitude of the surface cover type, vegetation coverage area, and water body distribution area includes: Extracting the surface cover type area ratio parameter, the vegetation cover area continuous distribution length parameter, and the water body distribution area shape complexity parameter at the previous time node in the spatial feature time series; Extracting the surface cover type area ratio parameter, the vegetation cover area continuous distribution length parameter, and the water body distribution area shape complexity parameter at the next time node in the spatial feature time series; Calculate the difference between the area ratio parameters of the land cover type at the next time node and the previous time node, determine the expansion direction or contraction direction of the land cover type according to the positive or negative sign of the difference, and determine the change amplitude of the land cover type according to the absolute value of the difference; Calculate the difference between the continuous distribution length parameter of the vegetation coverage area at the next time node and the previous time node, determine the direction of connectivity enhancement or weakening of the vegetation coverage area according to the positive or negative sign of the difference, and determine the change amplitude of the vegetation coverage area according to the absolute value of the difference; Calculate the difference in shape complexity parameters of the water distribution area between the next time node and the previous time node, determine the shape regularization direction or irregularity direction of the water distribution area according to the positive or negative sign of the difference, and determine the change amplitude of the water distribution area according to the absolute value of the difference.

8. The method for analyzing the evolution of watershed ecological environment by combining remote sensing images according to claim 6, characterized in that: The ecological state transition matrix is ​​constructed according to the change direction, change amplitude and heterogeneity distribution law. The ecological state transition matrix is ​​used to represent the ecological state transition probability of the target watershed from one time node to the next time node, including: defining an ecological state set of the target watershed, the ecological state set comprising a surface cover stable state, a surface cover expansion state, a surface cover contraction state, a vegetation connectivity stable state, a vegetation connectivity enhancement state, a vegetation connectivity weakening state, a water body shape regular state, and a water body shape irregular state; Counting the frequency of occurrence of the change direction and change magnitude at historical time nodes, and calculating the number of transitions from each initial ecological state to each target ecological state; Dividing the number of transitions by the total number of occurrences of the initial ecological state to obtain a transition probability from the initial ecological state to the target ecological state; Arrange the transition probabilities in row-priority order to generate a two-dimensional matrix structure with row indices representing initial ecological states and column indices representing target ecological states; The two-dimensional matrix structure is standardized to obtain the ecological state transfer matrix.

9. The method for analyzing the evolution of watershed ecological environment by combining remote sensing images according to claim 1, characterized in that: The ecological environment evolution trend of the target watershed is determined based on the ecological state transition characteristics, and the ecological environment evolution trend includes a stability evolution direction, a degradation evolution direction, and a restoration evolution direction, including: Extracting the surface cover type conversion path in the ecological state transition characteristics, analyzing the proportion of the surface cover type converted to the first ecological value type and the proportion of the surface cover type converted to the second ecological value type, where the ecological value of the first ecological value type is greater than the ecological value of the second ecological value type; Extracting the vegetation coverage change gradient in the ecological state transition characteristics, and analyzing the area proportion of regions where the vegetation coverage shows a positive change and the area proportion of regions where the vegetation coverage shows a negative change; Extracting the expansion and contraction trajectory of the water body distribution range in the ecological state transition characteristics, and analyzing the proportion of the length of the area showing an expansion trend and the proportion of the length of the area showing a contraction trend; Calculate the difference between the conversion ratio of the first ecological value type and the conversion ratio of the second ecological value type as an indicator of the evolution trend of the surface cover ecological environment; The difference between the area proportion of positive change regions and the area proportion of negative change regions was calculated as an indicator of the evolution trend of vegetation cover ecological environment. The difference between the length of the water body expansion area and the length of the water body contraction area is calculated as an indicator of the ecological environment evolution trend of the water body distribution; Based on the comprehensive evaluation results of the surface cover ecological environment evolution trend indicators, vegetation cover ecological environment evolution trend indicators and water body distribution ecological environment evolution trend indicators, the ecological environment evolution trend of the target watershed is determined to be a stable evolution direction, a degenerative evolution direction or a restorative evolution direction.

10. A watershed ecological environment evolution analysis system combined with remote sensing images, characterized in that: It includes a processor and a memory, the memory is connected to the processor, the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to implement the watershed ecological environment evolution analysis method combined with remote sensing images as described in any one of claims 1 to 9.

Images