Method for automatically observing water-holding capacity of litter in forest land

Abstract

The invention discloses a forest land litter water-holding capacity automatic observation method, which comprises the following steps: identifying topographic factor distribution of different slope gradients and altitudes, calculating vegetation structure parameters by combining vegetation indexes, dividing forest land environment heterogeneity regions according to the topographic factor distribution and the vegetation structure parameters, and generating a spatial distribution map layer, generating an observation point position layout scheme with regional representativeness, deploying a sensing device according to the observation point position layout scheme, and obtaining the dynamic change moment and the change amplitude of the litter water-holding capacity; and decomposing the dynamic change moment and the change amplitude of the litter water-holding capacity, generating continuous spatial distribution data of the forest land litter water-holding capacity, constructing a water-holding capacity prediction model, and calculating litter water-holding capacity observation values of different areas of the forest land in real time according to parameter coefficients of the prediction model. According to the invention, accurate observation of the water holding characteristic of the litter in the forest land is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring, and in particular to a method for automatically observing the water holding capacity of fallen leaves in forest land. Background Art

[0002] The litter layer in forest ecosystems is a crucial component of the hydrological cycle, and its water-holding capacity directly impacts watershed water conservation, soil erosion control, and ecosystem stability. Accurately monitoring changes in forest litter water holding capacity is crucial for forest hydrological research, ecological protection, and water resource management.

[0003] Traditional monitoring of litter water holding capacity relies primarily on periodic manual sampling and laboratory analysis. This method is not only time-consuming and labor-intensive, but also struggles to capture dynamic changes in water holding capacity. Manual monitoring is limited in frequency and cannot capture real-time variations influenced by environmental factors such as rainfall and evaporation. Furthermore, sampling sites are often poorly selected, making them less representative of the overall characteristics of complex forest environments.

[0004] The high heterogeneity of forest environments poses a fundamental challenge to water-holding capacity observation. Forests with different aspects, gradients, and altitudes exhibit distinct hydrological characteristics, and differences in stand age structure and tree species composition further exacerbate this spatial variability. This complex spatial heterogeneity directly leads to scientific challenges in the placement of observation points. Single or random point placement cannot accurately reflect the true distribution pattern of water-holding properties of litter within a region. Irrational placement of observation points further restricts the effective deployment of automated monitoring systems, as the lack of a representative observation network makes it difficult to obtain continuous, regionally representative observation data, even with advanced sensor technology.

[0005] Therefore, how to scientifically and rationally arrange observation points in a complex and changeable forest environment, build an automated observation network that can comprehensively reflect the water holding characteristics of litter in different forest types, and realize the continuous and accurate collection of water holding data has become a key issue in the development of automatic observation technology for water holding capacity of forest litter. Summary of the Invention

[0006] In order to solve the technical problems existing in the above-mentioned prior art, the present invention proposes an automatic observation method for the water holding capacity of forest litter, so as to realize the accurate observation of the water holding characteristics of forest litter.

[0007] To achieve the above-mentioned object, the present invention provides a method for automatically observing the water holding capacity of forest litter, comprising:

[0008] Conduct aerial scanning of the target forest land to identify the distribution of terrain factors at different slopes, gradients, and altitudes, and calculate vegetation structure parameters in combination with vegetation indices. Based on the distribution of terrain factors and vegetation structure parameters, divide the forest land environment heterogeneity areas and generate a spatial distribution layer.

[0009] Based on the area weights and hydrological characteristic differences of the heterogeneous regions in the spatial distribution layer, a regionally representative observation point layout plan is generated. Sensor devices are deployed according to the observation point layout plan to collect soil moisture data and soil weight data, respectively, to obtain the dynamic change trend and change amplitude of the water holding capacity of the litter;

[0010] Decomposing the dynamic change time and change amplitude of the water holding capacity of the litter to obtain the water holding characteristic change pattern at different time scales, and extending the water holding characteristic change pattern at different time scales to the entire forest area to generate continuous spatial distribution data of the water holding capacity of the forest litter;

[0011] The quantitative relationship between the water holding capacity of litter and meteorological data is obtained based on the continuous spatial distribution data, a water holding capacity prediction model is constructed, and the observed water holding capacity of litter in different areas of the forest is calculated in real time based on the parameter coefficients of the prediction model.

[0012] Compared with the prior art, the present invention has the following advantages and technical effects:

[0013] The present invention discloses a method for automatically observing the water holding capacity of forest litter. The method uses multispectral scanning by unmanned aerial vehicles to obtain terrain and vegetation parameters, divides the forest into heterogeneous areas and arranges observation points, and deploys a network of soil moisture and weighing sensors to realize automatic monitoring. Kalman filtering and time series analysis are used to process monitoring data, extract water holding capacity variation characteristics, and use spatial interpolation to generate continuous distribution data. A water holding capacity prediction model is established in combination with meteorological data to realize dynamic prediction and model optimization of water holding capacity of litter. The present invention realizes the precise observation and prediction of water holding characteristics of forest litter, provides an effective tool for forest hydrological process research and water resource management, can improve the water regulation capacity of forest land, and promote the sustainable development of forest ecosystems. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0015] Figure 1 This is a flow chart of a method for automatically observing water holding capacity of forest litter according to an embodiment of the present invention. DETAILED DESCRIPTION

[0016] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0017] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0018] This embodiment proposes a method for automatically observing the water holding capacity of forest litter. Figure 1 ,include:

[0019] Conduct aerial scanning of the target forest land to identify the distribution of terrain factors at different slopes, gradients, and altitudes, and calculate vegetation structure parameters in combination with vegetation indices. Based on the distribution of terrain factors and vegetation structure parameters, divide the forest land environment heterogeneity areas and generate a spatial distribution layer.

[0020] Based on the area weights and hydrological characteristic differences of the heterogeneous regions in the spatial distribution layer, a regionally representative observation point layout plan is generated. Sensor devices are deployed according to the observation point layout plan to collect soil moisture data and soil weight data, respectively, to obtain the dynamic change trend and change amplitude of the water holding capacity of the litter;

[0021] Decomposing the dynamic change time and change amplitude of the water holding capacity of the litter to obtain the water holding characteristic change pattern at different time scales, and extending the water holding characteristic change pattern at different time scales to the entire forest area to generate continuous spatial distribution data of the water holding capacity of the forest litter;

[0022] The quantitative relationship between the water holding capacity of litter and meteorological data is obtained based on the continuous spatial distribution data, a water holding capacity prediction model is constructed, and the observed water holding capacity of litter in different areas of the forest is calculated in real time based on the parameter coefficients of the prediction model.

[0023] Furthermore, aerial scanning of the target forest is carried out to identify the distribution of terrain factors at different slopes, gradients and altitudes, and vegetation structure parameters are calculated in combination with vegetation indices, including:

[0024] A multispectral sensor is used to carry out aerial photography of the target forest land, obtain multispectral image data, and generate a spectral reflectance data set;

[0025] Calculating the vegetation index using the normalized difference vegetation index formula using the spectral reflectance data set to obtain a vegetation index distribution;

[0026] Combined with the vegetation index distribution, a random forest algorithm is used to classify the forest age structure and tree species composition to obtain the vegetation structure parameters.

[0027] Specifically, a multispectral sensor equipped with an unmanned aerial vehicle (UAV) was used to take aerial photos of the forest to obtain multispectral image data and generate a spectral reflectance dataset. Using the spectral reflectance dataset, the vegetation index was calculated using the normalized difference vegetation index formula NDVI = (NIR-RED) / (NIR+RED), where NIR is the reflectance of the near-infrared band and RED is the reflectance of the red band, to obtain the vegetation index distribution. If the NDVI value of a certain area in the vegetation index distribution is higher than the preset normalization threshold, the spectral reflectance data is classified using the support vector machine algorithm to determine the aspect distribution, slope distribution, and altitude distribution, thereby obtaining the terrain factor distribution. Based on the terrain factor distribution, the K-means clustering algorithm was applied to perform cluster analysis on the aspect, slope, and altitude data, dividing the environmental heterogeneity areas and obtaining the heterogeneous regional distribution. Based on the heterogeneous regional distribution and combined with the vegetation index distribution, the random forest algorithm was used to classify the forest age structure and tree species composition to obtain the vegetation structure parameters.

[0028] Furthermore, the forest environment heterogeneity regions are divided according to the distribution of terrain factors and vegetation structure parameters, and a spatial distribution layer is generated, including:

[0029] If the forest age structure in the vegetation structure parameter matches the preset forest age threshold, the slope, aspect and altitude are extracted from the terrain factor data to obtain the terrain factor distribution;

[0030] According to the distribution of the terrain factors, the distribution of the terrain factors and the vegetation structure parameters are spatially superimposed and analyzed using geographic information system technology to generate a preliminary spatial distribution layer;

[0031] Performing format conversion on the preliminary spatial distribution layer through rasterization processing technology to obtain a distribution layer in vector format;

[0032] If there is an area in the distribution layer in the vector format whose spatial resolution is lower than the preset resolution threshold, an interpolation algorithm is used to perform spatial interpolation processing on the low-resolution area to obtain an optimized spatial distribution layer;

[0033] Based on the optimized spatial distribution layer, vector analysis technology is used to extract the boundaries of the environmental heterogeneity area to generate the final spatial distribution layer.

[0034] Specifically, the preliminary spatial distribution layer is converted into a vector format through rasterization technology. Rasterization first divides the layer data into regular grid cells, such as a 10m x 10m grid, and then converts the attribute values ​​of each grid into a vector format to generate the final spatial distribution layer.

[0035] For example, the converted vector layer clearly shows the distribution of mature coniferous forests in a high-altitude area, which accounts for 15% of the total forest area. This provides a spatial basis for precise forest resource management. Such vector layers facilitate subsequent analysis and visualization, improving data utilization efficiency.

[0036] The application of these methods can significantly improve the accuracy and efficiency of forestland monitoring. UAV multispectral imagery provides a high-resolution data source for vegetation index calculation. Support vector machines and K-means clustering effectively delineate regions of topographic and environmental heterogeneity. Random forests further refine vegetation structure classification. Geographic Information Systems and rasterization enable spatial visualization of data. These mutually reinforcing technologies form a comprehensive approach to forestland ecological assessment, providing a scientific basis for forest conservation, resource management, and ecological planning.

[0037] Furthermore, a regionally representative observation point layout plan is generated, including:

[0038] Through the spatial distribution layer, the area weight and hydrological characteristic data of each heterogeneous area are obtained, the area proportion and the coefficient of variation of the hydrological characteristics of each area are calculated, and the regional characteristic data set is obtained;

[0039] Based on the regional characteristic dataset, a stratified sampling algorithm is used to calculate the number of observation points in each heterogeneous area based on the area weight and the coefficient of variation, and the distribution of the number of observation points is obtained. A grid division technique is then applied to divide each heterogeneous area into a uniform grid, and the spatial coordinates of the observation points are generated within the grid based on the point density to obtain a point coordinate dataset.

[0040] Based on the point coordinate data set and the coefficient of variation of the hydrological characteristics, the point coordinates are classified, the representative weight of each point is determined, and the representative distribution of the point is obtained;

[0041] Through the representative distribution of the points, the point coordinates are superimposed with the spatial distribution of the heterogeneous area to generate a spatial distribution layer of the observation points and the observation point layout plan.

[0042] Specifically, using the spatial distribution layer, the area weight and hydrological characteristic data of each heterogeneous region are obtained. The area share and coefficient of variation of each region's hydrological characteristics are calculated to obtain a regional characteristic dataset. Based on the regional characteristic dataset, a stratified sampling algorithm is used to calculate the number of observation points within each heterogeneous region based on the area weight and coefficient of variation, resulting in a point quantity distribution. If the area weight of a region in the point quantity distribution exceeds a preset threshold and the coefficient of variation exceeds a preset value, the increased point density is calculated using the density adjustment formula D = k(AV), where D is the point density, A is the area weight, V is the coefficient of variation, and k is a constant coefficient. This results in an adjusted point density distribution. Using this adjusted point density distribution, a gridding technique is applied to divide each heterogeneous region into a uniform grid. The spatial coordinates of the observation points within the grid are generated based on the point density, resulting in a point coordinate dataset. Based on this point coordinate dataset and the coefficient of variation of the hydrological characteristics, a random forest algorithm is used to classify the point coordinates, determine the representative weight of each point, and obtain a point representativeness distribution. Through the representative distribution of points and the application of geographic information system technology, the point coordinates are superimposed on the spatial distribution of heterogeneous areas to generate a spatial distribution layer of observation points and obtain the final observation point layout plan.

[0043] In one embodiment, a random forest algorithm is used to classify points based on a dataset of point coordinates and the coefficient of variation of hydrological characteristics, determining representative weights. For example, points in high-altitude, steep slopes have a higher weight of 0.7 due to the large variation in hydrological characteristics; points in low-altitude, gentle slopes have a weight of 0.5. This classification method prioritizes monitoring in key areas. The random forest algorithm uses multiple decision trees to analyze the relationship between point coordinates and hydrological characteristics, ensuring reliable classification results.

[0044] Using geographic information system technology, the coordinates of the observation points are overlaid with the heterogeneous regions to generate a spatial distribution layer of the observation points. For example, 12 observation points in the high-altitude steep slope area are concentrated in areas with slopes greater than 30 degrees, while 20 observation points in the low-altitude gentle slope area are distributed across flat areas. The resulting layer visually displays the correlation between the observation points and topographic and hydrological characteristics, facilitating subsequent monitoring planning. This layout optimizes the spatial representativeness of the observation points and improves monitoring efficiency.

[0045] Furthermore, sensor devices are deployed according to the observation point layout plan to collect soil moisture data and soil weight data respectively, including:

[0046] Extract terrain factors and vegetation structure parameters from the observation point layout plan, use K-means clustering algorithm to classify the points, and obtain the classified observation point set;

[0047] Based on the classified observation point set, combined with terrain factors and vegetation structure parameters, the weighted average method is used to calculate the data collection priority of each point and determine the differentiated collection frequency plan;

[0048] The sensor network is configured through the LoRa wireless communication protocol, the acquisition frequency parameters of each point are obtained from the differentiated acquisition frequency scheme, the acquisition frequencies of the soil moisture sensor and the weighing sensor are set, and the configured sensor network is obtained;

[0049] Soil moisture and soil weight data are collected from the configured sensor network. If the data completeness is lower than the preset completeness threshold, the missing data are supplemented by linear interpolation to obtain a complete dataset.

[0050] Specifically, terrain factor data is obtained through remote sensing images and digital elevation models, and vegetation structure parameters are extracted in combination with the vegetation cover index to generate an observation point distribution map. Terrain factors and vegetation structure parameters are extracted from the observation point distribution map, and the points are classified using the K-means clustering algorithm to obtain a classified observation point set. Based on the classified observation point set, the weighted average method is used to calculate the data collection priority of each point in combination with the terrain factors and vegetation structure parameters, and a differentiated collection frequency plan is determined. The sensor network is configured through the LoRa wireless communication protocol, and the collection frequency parameters of each point are obtained from the differentiated collection frequency plan. The collection frequency of the soil moisture sensor and weighing sensor is set to obtain the configured sensor network. Soil moisture and soil weight data are collected from the configured sensor network. If the data integrity is lower than the preset integrity threshold, the missing data is supplemented by linear interpolation to obtain a complete data set.

[0051] Furthermore, the dynamic change time and change range of the water holding capacity of the litter are obtained, including:

[0052] The soil moisture data and the soil weight data are smoothed in real time using a Kalman filter algorithm, and multi-sensor data are fused using a state estimation method to obtain smoothed time series data;

[0053] The smoothed time series data is processed using a sliding window method, the data gradient within each time window is calculated, and a dynamic change sequence of the data gradient is obtained;

[0054] If a gradient value in the dynamic change sequence of the data gradient exceeds the preset gradient threshold, it is judged as a rainfall event or evaporation event, and the time and type of the triggering event are determined;

[0055] Extracting soil moisture data and weight data for a corresponding time period from the time series data according to the time and type of the triggering event, and calculating the change in water holding capacity of the litter;

[0056] The trend of the change amplitude of the water holding capacity of the litter is extracted by a time series analysis method to obtain the dynamic change trend of the water holding capacity of the litter.

[0057] Specifically, in one possible implementation, the Kalman filter algorithm, when used to smooth soil moisture and soil weight data in real time, can be considered a state estimation tool. Through two steps, prediction and updating, it combines sensor observations with the system model to reduce noise interference. For example, suppose a soil moisture sensor in a forest collects data every 10 minutes. The raw data may fluctuate significantly due to environmental noise, such as humidity values ​​fluctuating erratically between 50% and 55%. The Kalman filter uses historical data to predict the current state and then, combined with new observations, corrects it to generate a smoothed sequence of humidity values, such as one that stabilizes around 52%. This smoothing process more accurately reflects the true changes in soil moisture, facilitating subsequent analysis.

[0058] Specifically, when fusing multi-sensor data, the Kalman filter can integrate the outputs of soil moisture sensors and load cells. For example, the soil moisture sensor measures 53% humidity, while the weight change recorded by the load cell suggests possible rainfall. Using state estimation, the algorithm fuses these two data types to generate consistent time series data, such as hourly soil moisture content and litter weight change series. This fusion provides a comprehensive picture of forest moisture dynamics and improves data reliability.

[0059] Furthermore, the dynamic change time and amplitude of water holding capacity of litter were decomposed to obtain the change patterns of water holding characteristics at different time scales, including:

[0060] The water holding capacity data were processed by time series decomposition algorithm to extract trend component, seasonal component and random component, and obtain the time series characteristics of each component.

[0061] The trend component is analyzed using an autocorrelation function, a periodic parameter is calculated, and a long-term variation pattern of the trend component is determined;

[0062] For the seasonal component, applying Fourier transform to extract periodic frequency characteristics to obtain the periodic pattern of the seasonal component;

[0063] If the autocorrelation function value of the random component is lower than a preset autocorrelation function threshold, the random component is smoothed by a moving average method to obtain a smoothed random component sequence;

[0064] According to the smoothed random component sequence, combined with the periodic patterns of the trend component and the seasonal component, the water holding characteristic variation patterns under the different time scales are obtained.

[0065] Specifically, a seasonal decomposition algorithm was used to process the water holding capacity time series data, extracting trend components, seasonal components, and random components to obtain sequences of each component. For the extracted trend component, an autoregressive model was applied to analyze its time series characteristics and determine the long-term variation pattern of the trend component. Based on the seasonal component, the periodic frequency was calculated using fast Fourier transform to obtain the periodic regularity of the seasonal component. If the autocorrelation function value of the random component is lower than the preset threshold, the random component is processed using exponential smoothing to obtain a smoothed random component sequence; if it is higher than the preset threshold, the original random component sequence is retained. By combining the long-term variation pattern of the trend component, the periodic regularity of the seasonal component, and the smoothed random component sequence, the variation pattern of water holding characteristics at different time scales was obtained. The water holding capacity variation pattern was reduced in dimensionality using principal component analysis, the main eigenvectors were extracted, and the key driving factors of water holding capacity changes were determined.

[0066] Furthermore, continuous spatial distribution data of forest litter water holding capacity is generated, including:

[0067] The spatial interpolation algorithm is used to extend the water holding characteristic change pattern of each observation point to the entire forest area, and the water holding capacity estimation of unobserved locations is calculated using the inverse distance weighted method. The interpolation weight coefficient is adjusted according to the similarity of terrain factors and vegetation structure to generate continuous spatial distribution data of forest litter water holding capacity.

[0068] Specifically, when obtaining litter water holding capacity data, terrain factor data, and vegetation structure data at each observation point within the forest, an initial data set can be constructed by combining field sampling and remote sensing technology. Assume that 100 observation points are selected in a certain forest, and each point records data such as litter water holding capacity, terrain slope, slope aspect, vegetation coverage, and tree species distribution. Water holding capacity data is measured by weighing, terrain factor data is extracted using a digital elevation model, and vegetation structure data is based on drone aerial image analysis. The initial data set must ensure that the points are evenly distributed and cover different terrain and vegetation types in the forest to ensure the representativeness of the data.

[0069] In one possible implementation, the inverse distance weighted (IDW) algorithm can be used to estimate water holding capacity at unobserved locations by spatially interpolating the water holding capacity data at observed locations. For example, an unobserved point is 50 meters, 100 meters, and 150 meters from three observed points, with water holding capacities of 200 mm, 180 mm, and 160 mm, respectively. Using a weighted calculation using the inverse of distance, a preliminary estimate of the water holding capacity at this point is approximately 185 mm. This method uses the inverse of spatial distance as a weight, so closer points contribute more to the estimated value, making it suitable for scenarios with relatively uniform spatial distribution.

[0070] Specifically, when calculating the terrain similarity coefficient, the terrain difference between the unobserved point and the observed point can be quantified by the weighted Euclidean distance formula based on factors such as slope, aspect, and altitude. For example, the slope of an unobserved point is 5 degrees and the aspect is north, while the slope of the observed point is 6 degrees and the aspect is northeast. The terrain similarity coefficient between the two can be calculated by the standardized factor difference, and the value obtained is 0.85. Similarly, the vegetation similarity coefficient can be calculated based on the degree of matching between coverage and tree species. For example, the coverage of the unobserved point is 70%, the main tree species is pine, and the coverage of the observed point is 65%. The tree species are the same, and the similarity coefficient can be set to 0.90. These coefficients are used to adjust the weights of the inverse distance weighted algorithm to improve the estimation accuracy.

[0071] Furthermore, a water holding capacity prediction model is constructed, including:

[0072] Obtain continuous spatial distribution data of water holding capacity of litter within the forest area and meteorological data to construct an initial data set, wherein the meteorological data includes rainfall, evaporation, temperature and humidity meteorological data;

[0073] Using the initial data set, a stepwise regression method is used to analyze the correlation between meteorological data and water holding capacity, screen out significant influencing factors, and obtain a significant influencing factor set;

[0074] Using multiple regression analysis, a quantitative relationship between water holding capacity and meteorological factors is established for a set of significant influencing factors, and parameter coefficients are determined. The predicted water holding capacity at each location is calculated using the parameter coefficients in combination with continuous spatial distribution data to obtain a predicted value distribution. If the deviation between the predicted value distribution and the actual observed value exceeds a prediction threshold, the parameter coefficients are adjusted through cross-validation to obtain the optimized parameter coefficients.

[0075] The optimized parameter coefficients are used to rerun the multiple regression analysis, update the distribution of water holding capacity prediction values, and obtain an adjusted prediction distribution. Based on the adjusted prediction distribution and combined with the spatiotemporal variation characteristics of meteorological data, the water holding capacity prediction model is generated.

[0076] Building a water holding capacity prediction model also includes:

[0077] Obtain measured data on the water holding capacity of litter in each forest area, and use sensors to collect environmental variables to obtain an initial data set;

[0078] The random forest algorithm was applied to calculate the predicted water holding capacity of litter in each area using the initial data set and the parameters of the water holding capacity prediction model.

[0079] Based on the deviation between the predicted value and the measured data, the prediction error is calculated, and the root mean square error formula is used to obtain the error quantification result. If the error quantification result exceeds the preset quantification threshold, the error distribution characteristics are extracted, and the parameters of the water holding capacity prediction model are optimized using the gradient descent method to obtain the updated model parameters;

[0080] Based on the updated model parameters, the predicted water holding capacity of litter in each region is recalculated to obtain a new prediction data set.

[0081] The model convergence is judged by comparing the new prediction data set with the measured data. If the error still exceeds the preset threshold, the optimization steps are repeated until the final optimized model parameters are obtained.

[0082] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for automatically observing the water holding capacity of forest litter, characterized in that: include: Conduct aerial scanning of the target forest land to identify the distribution of terrain factors at different slopes, gradients, and altitudes, and calculate vegetation structure parameters in combination with vegetation indices. Based on the distribution of terrain factors and vegetation structure parameters, divide the forest land environment heterogeneity areas and generate a spatial distribution layer. Based on the area weights and hydrological characteristic differences of the heterogeneous regions in the spatial distribution layer, a regionally representative observation point layout plan is generated. Sensor devices are deployed according to the observation point layout plan to collect soil moisture data and soil weight data, respectively, to obtain the dynamic change trend and change amplitude of the water holding capacity of the litter; Decomposing the dynamic change time and change amplitude of the water holding capacity of the litter to obtain the water holding characteristic change pattern at different time scales, and extending the water holding characteristic change pattern at different time scales to the entire forest area to generate continuous spatial distribution data of the water holding capacity of the forest litter; The quantitative relationship between the water holding capacity of litter and meteorological data is obtained based on the continuous spatial distribution data, a water holding capacity prediction model is constructed, and the observed water holding capacity of litter in different areas of the forest is calculated in real time based on the parameter coefficients of the prediction model.

2. The method for automatically observing the water holding capacity of forest litter according to claim 1, characterized in that: Conduct aerial scanning of the target forest land to identify the distribution of terrain factors at different slopes, gradients, and altitudes, and calculate vegetation structure parameters based on vegetation indices, including: A multispectral sensor is used to carry out aerial photography of the target forest land, obtain multispectral image data, and generate a spectral reflectance data set; Calculating the vegetation index using the normalized difference vegetation index formula using the spectral reflectance data set to obtain a vegetation index distribution; Combined with the vegetation index distribution, a random forest algorithm is used to classify the forest age structure and tree species composition to obtain the vegetation structure parameters.

3. The method for automatically observing the water holding capacity of forest litter according to claim 2, characterized in that: The forest environment heterogeneity regions are divided according to the distribution of the terrain factors and the vegetation structure parameters, and a spatial distribution layer is generated, including: If the forest age structure in the vegetation structure parameter matches the preset forest age threshold, the slope, aspect and altitude are extracted from the terrain factor data to obtain the terrain factor distribution; According to the distribution of the terrain factors, the distribution of the terrain factors and the vegetation structure parameters are spatially superimposed and analyzed using geographic information system technology to generate a preliminary spatial distribution layer; Performing format conversion on the preliminary spatial distribution layer through rasterization processing technology to obtain a distribution layer in vector format; If there is an area in the distribution layer in the vector format whose spatial resolution is lower than the preset resolution threshold, an interpolation algorithm is used to perform spatial interpolation processing on the low-resolution area to obtain an optimized spatial distribution layer; Based on the optimized spatial distribution layer, vector analysis technology is used to extract the boundaries of the environmental heterogeneity area to generate the final spatial distribution layer.

4. The method for automatically observing the water holding capacity of forest litter according to claim 1, characterized in that: Generate a regionally representative observation point layout plan, including: Through the spatial distribution layer, the area weight and hydrological characteristic data of each heterogeneous area are obtained, the area proportion and the coefficient of variation of the hydrological characteristics of each area are calculated, and the regional characteristic data set is obtained; Based on the regional characteristic dataset, a stratified sampling algorithm is used to calculate the number of observation points in each heterogeneous area based on the area weight and the coefficient of variation, and the distribution of the number of observation points is obtained. A grid division technique is then applied to divide each heterogeneous area into a uniform grid, and the spatial coordinates of the observation points are generated within the grid based on the point density to obtain a point coordinate dataset. Based on the point coordinate data set and the coefficient of variation of the hydrological characteristics, the point coordinates are classified, the representative weight of each point is determined, and the representative distribution of the point is obtained; Through the representative distribution of the points, the point coordinates are superimposed with the spatial distribution of the heterogeneous area to generate a spatial distribution layer of the observation points and the observation point layout plan.

5. The method for automatically observing the water holding capacity of forest litter according to claim 4, characterized in that: Deploy sensor devices according to the observation point layout plan to collect soil moisture data and soil weight data respectively, including: Extracting terrain factors and vegetation structure parameters from the observation point layout plan, classifying the points using a K-means clustering algorithm, and obtaining a classified observation point set; Based on the classified observation point set, combined with terrain factors and vegetation structure parameters, a weighted average method is used to calculate the data collection priority of each point and determine a differentiated collection frequency plan; The sensor network is configured through the LoRa wireless communication protocol, the acquisition frequency parameters of each point are obtained from the differentiated acquisition frequency scheme, the acquisition frequencies of the soil moisture sensor and the weighing sensor are set, and the configured sensor network is obtained; Soil moisture and soil weight data are collected from the configured sensor network. If the data completeness is lower than the preset completeness threshold, the missing data are supplemented by linear interpolation to obtain a complete dataset.

6. The method for automatically observing the water holding capacity of forest litter according to claim 5, characterized in that: Obtaining the dynamic change time and change range of the water holding capacity of the litter, including: The soil moisture data and the soil weight data are smoothed in real time using a Kalman filter algorithm, and multi-sensor data are fused using a state estimation method to obtain smoothed time series data; The smoothed time series data is processed using a sliding window method, the data gradient within each time window is calculated, and a dynamic change sequence of the data gradient is obtained; If a gradient value in the dynamic change sequence of the data gradient exceeds the preset gradient threshold, it is judged as a rainfall event or evaporation event, and the time and type of the triggering event are determined; According to the time and type of the triggering event, soil moisture data and weight data of the corresponding time period are extracted from the time series data, and the change range of the water holding capacity of the litter is calculated; The trend of the change amplitude of the water holding capacity of the litter is extracted by a time series analysis method to obtain the dynamic change trend of the water holding capacity of the litter.

7. The method for automatically observing the water holding capacity of forest litter according to claim 1, characterized in that: The dynamic change time and change amplitude of the water holding capacity of the litter are decomposed to obtain the change patterns of water holding characteristics at different time scales, including: The water holding capacity data were processed by time series decomposition algorithm to extract trend component, seasonal component and random component, and obtain the time series characteristics of each component. The autocorrelation function is used to analyze the trend components, calculate the periodic parameters, and determine the long-term change pattern of the trend components; For the seasonal component, applying Fourier transform to extract periodic frequency characteristics to obtain the periodic pattern of the seasonal component; If the autocorrelation function value of the random component is lower than a preset autocorrelation function threshold, the random component is smoothed by a moving average method to obtain a smoothed random component sequence; According to the smoothed random component sequence, combined with the periodic patterns of the trend component and the seasonal component, the water holding characteristic variation patterns under the different time scales are obtained.

8. The method for automatically observing water holding capacity of forest litter according to claim 7, characterized in that: Generate continuous spatial distribution data of forest litter water holding capacity, including: A spatial interpolation algorithm was used to extend the water holding characteristic variation pattern of each observation point to the entire forest area. The water holding capacity estimation of unobserved locations was calculated using the inverse distance weighted method. The interpolation weight coefficient was adjusted according to the similarity of terrain factors and vegetation structure to generate continuous spatial distribution data of forest litter water holding capacity.

9. The method for automatically observing water holding capacity of forest litter according to claim 1, characterized in that: Constructing the water holding capacity prediction model includes: Obtain continuous spatial distribution data of water holding capacity of litter within the forest area and meteorological data to construct an initial data set, wherein the meteorological data includes rainfall, evaporation, temperature and humidity meteorological data; Using the initial data set, a stepwise regression method is used to analyze the correlation between meteorological data and water holding capacity, screen out significant influencing factors, and obtain a significant influencing factor set; Using multiple regression analysis, a quantitative relationship between water holding capacity and meteorological factors is established for a set of significant influencing factors, and parameter coefficients are determined. The predicted water holding capacity at each location is calculated using the parameter coefficients in combination with continuous spatial distribution data to obtain a predicted value distribution. If the deviation between the predicted value distribution and the actual observed value exceeds a prediction threshold, the parameter coefficients are adjusted through cross-validation to obtain the optimized parameter coefficients. The optimized parameter coefficients are used to rerun the multiple regression analysis, update the distribution of water holding capacity prediction values, and obtain an adjusted prediction distribution. Based on the adjusted prediction distribution and combined with the spatiotemporal variation characteristics of meteorological data, the water holding capacity prediction model is generated.

10. The method for automatically observing water holding capacity of forest litter according to claim 9, characterized in that: Constructing the water holding capacity prediction model also includes: Obtain measured data on the water holding capacity of litter in each forest area, and use sensors to collect environmental variables to obtain an initial data set; The random forest algorithm was applied to calculate the predicted water holding capacity of litter in each area using the initial data set and the parameters of the water holding capacity prediction model. Based on the deviation between the predicted value and the measured data, the prediction error is calculated, and the root mean square error formula is used to obtain the error quantification result. If the error quantification result exceeds the preset quantification threshold, the error distribution characteristics are extracted, and the parameters of the water holding capacity prediction model are optimized using the gradient descent method to obtain the updated model parameters; Based on the updated model parameters, the predicted water holding capacity of litter in each region is recalculated to obtain a new prediction data set. The model convergence is judged by comparing the new predicted data set with the measured data. If the error still exceeds the preset threshold, the optimization steps are repeated until the final optimized model parameters are obtained.

Images