Desert riparian forest biomass estimation method, device, equipment and medium

Abstract

The invention discloses a desert riparian forest biomass estimation method and device, equipment and a medium, and relates to the technical field of biomass monitoring, and the method comprises the steps: firstly, researching tree wheel information at different trunk heights, and obtaining the annual variation of the annual biomass of trunks through the radial growth; based on the three-dimensional scanning information of the trunk, constructing a trunk hierarchical volume algorithm to calculate the volume of the trunk, and establishing a trunk volume model to calculate the biomass of the trunk; and finally, based on field investigation statistical analysis, analyzing a trunk and branch biomass distribution pattern. According to the process, trunk morphological characteristics and growth characteristics represented by tree wheel information and three-dimensional scanning information are introduced into estimation of the trunk volume, so that the problem of large trunk biomass estimation difficulty caused by large individual differences of populus euphratica and irregular tree distribution is solved, and the biomass estimation difficulty of the trunk biomass is improved by combining the trunk and branch biomass distribution rule of a riparian forest. The populus euphratica individual difference phenomenon is further considered, so that the statistical accuracy of the large-area populus euphratica biomass is integrally improved.

Description

Technical Field

[0001] The present invention relates to the technical field of biomass monitoring, and in particular to a method, device, equipment and medium for estimating the biomass of desert riparian forests. Background Art

[0002] Desert riverbank basins are located in inland extremely arid areas, with the dual characteristics of rich natural resources and fragile ecological environment. There are a large number of Populus euphratica forests distributed on both sides of the basin. Desert riverbank forests composed of Populus euphratica are the main vegetation type in desert riverbank basins. They are natural tree communities close to the top succession in the process of basin vegetation succession under temperate desert climate conditions. They play an important role in resisting wind and sand, curbing desertification, maintaining regional ecological security, and protecting biodiversity. At the same time, desert riverbank forests, as the most productive part of desert ecosystems, play a vital role in maintaining ecological balance. Populus euphratica forest biomass is the basis for studying the functions and values of riverbank forest ecosystems, and is also an important indicator for evaluating ecosystem productivity. Forest biomass estimation has become crucial to forest resource management and protection.

[0003] Currently, forest biomass estimation in riparian forest ecosystems typically involves field measurements, remote sensing, and modeling. Field measurements, including direct harvesting, average standard tree, and stratified standard tree methods, offer high accuracy but require on-site tree felling and forest research, which is time-consuming and labor-intensive and has negative impacts on forest conservation. Remote sensing utilizes spatial information to analyze the spatial distribution of forest resources and, in combination with topographic information, explores the spatial distribution patterns of forests. However, remote sensing technology has regional limitations, poor portability, an incomplete theoretical foundation, a lack of understanding and knowledge of physical mechanisms, and a lack of logical relationships between parameters. Furthermore, remotely sensed vegetation indices are susceptible to contamination by snow / ice, atmospheric effects, and floating clouds, leading to significant deviations in the results. Furthermore, remote sensing satellite observations are typically instantaneous and have low temporal resolution. Consequently, remote sensing methods are significantly affected by meteorological factors and suffer from low accuracy. Modeling methods often use parameters such as diameter at breast height (DBH), tree height, and their derived factors to construct tree volume models or biomass models. These methods are widely used and require minimal damage to forests, making them relatively convenient and effective methods for determining forest biomass.

[0004] Based on the measurement of model method, some researchers proposed a method to reconstruct and render the volume of trees using tree photos. It usually requires modeling of different biomasses, but the number of modeling is huge and it is not suitable for statistics of large-scale biomass. In response to this phenomenon, researchers established a trunk biomass model to conduct statistics on forest biomass, but the individual shapes of Populus euphratica vary greatly and there is no regularity in their distribution. The current biomass research methods are relatively traditional and the applied parameters are single. It is difficult to take into account the differences in Populus euphratica tree shapes, resulting in low statistical accuracy for large-scale Populus euphratica biomass. Summary of the Invention

[0005] The embodiments of the present invention provide a method, device, equipment and medium for estimating the biomass of desert riparian forests, which can solve the problem in the prior art that the current methods are difficult to take into account the differences in the shapes of Populus euphratica trees, resulting in low statistical accuracy for the biomass of Populus euphratica over a large area.

[0006] An embodiment of the present invention provides a method for estimating the biomass of a desert riparian forest, comprising the following steps: Obtain the tree ring width and trunk density at different heights of riparian forest trees, and based on the tree ring width and trunk density, obtain the radial growth of trees in different time periods; Obtaining 3D point cloud data of multiple trees in a riparian forest, obtaining the volumes of the multiple trees based on the 3D point cloud data, and establishing a trunk volume model for each individual tree; wherein the trunk volume model for each individual tree characterizes the volumetric morphology of the multiple trees in the riparian forest; The trunk biomass of individual trees was obtained based on the trunk volume model and trunk density of individual trees. The ratio between the trunk volume and branch volume of multiple trees in riparian forests at different time periods was calculated, and the ratio between the trunk volume and branch volume of individual trees at different time periods was obtained. Based on the ratio between the trunk volume and branch volume of individual trees at different time periods and the trunk biomass of individual trees, the ratio between the trunk biomass and branch biomass of individual trees at different time periods was obtained. The biomass of individual trees was obtained based on the radial growth of trees in different time periods and the ratio of trunk biomass to branch biomass of individual trees in different time periods. Combined with the number of trees counted in the riparian forest, the estimated result of the riparian forest biomass was obtained.

[0007] Preferably, obtaining the radial growth amount of the tree in different time periods includes: Using the tree ring samples and trunk density samples collected from the riparian forest in the field, the density of trunk dry matter was obtained using the collected density samples; Extract the tree ring width and use the frustum formula to calculate the volume of the corresponding height segment of Populus euphratica. The difference between the volume in year i and the volume in year i-1 is: , the volume difference is the radial growth of the tree in year i.

[0008] Preferably, obtaining the volumes of the plurality of trees based on the three-dimensional point cloud data includes: Obtain 3D point cloud data of multiple riparian forest tree trunks, and segment the 3D point cloud data of the tree trunks into n layers in the vertical direction; The 3D point cloud data of a tree is divided into n slices in the vertical direction. In each slice, two upper and lower polygons are generated to determine the perimeter of the upper and lower planes of the polygon, and the corresponding points with the top and bottom surfaces are found in the vertical direction. Based on the periphery of the upper and lower planes of the determined polygon, as well as the corresponding points on the top and bottom surfaces, an area function of the polygonal area is generated. Based on the area function, an area integral function is constructed to calculate the volume of the slice. The volumes of n slices are obtained and superimposed to obtain the volume of the entire tree trunk.

[0009] Preferably, the area integral function is: ; Where: v represents the volume of the entire tree trunk; r represents the trunk radius of the smallest periphery of the slice; r(x) represents the trunk radius of the largest periphery of the slice; c(r i ) represents the i-th polygonal area.

[0010] Preferably, the construction of the trunk volume model of the single tree includes: Based on the 3D point cloud data, the diameter at breast height, base diameter, tree height and crown width of each riparian forest tree were obtained; Based on the trunk volume of each riparian forest tree and the corresponding diameter at breast height, base diameter, tree height, and crown width information of each riparian forest tree, linear regression analysis in Scikit-learn software was used to fit the model and obtain the trunk volume model; the trunk volume model is: ; Where: v is the volume; DBH is the diameter at breast height; BD is the base diameter of the trunk; H is the height of the tree; SH is the height of the trunk; DH 2 Derivative factors representing tree height and diameter at breast height.

[0011] Preferably, obtaining the ratio between the trunk biomass and the branch biomass of a single tree in different time periods includes: Field statistics were collected on the diameter at breast height, base diameter, tree height, and crown width of multiple riparian forest tree trunks. One part collected information on the length, base diameter, and number of primary and secondary branches in the riparian forests, while another part collected information on the number and base diameter of tertiary branches in the riparian forests. Furthermore, information on the trunk and branch combinations of multiple riparian forest trees was obtained. In statistics, the difference in the number of primary and secondary branches is related to the tree age and base diameter. The base diameter of primary and secondary branches is smaller, and the number of tertiary branches is larger. Trees with dense tertiary branches have a rounded crown shape, good growth, and dense crowns. Trees with sparse tertiary branches have an incomplete crown shape and poor growth. Based on the statistical number of primary, secondary and tertiary branches, and using the crown biomass model, trunk volume model and trunk density, the trunk and branch biomass distribution ratios and the trunk and crown biomass distribution ratios of individual trees in riparian forests in different time periods were obtained respectively; based on the trunk and branch biomass distribution ratios and the trunk and crown biomass distribution ratios, the ratios between trunk biomass and branch biomass of individual trees in different time periods were obtained.

[0012] An embodiment of the present invention further provides a device for estimating the biomass of a desert riparian forest, comprising: The tree ring module is used to obtain the tree ring width and trunk density of riparian forest trees at different heights, and to obtain the radial growth of trees in different time periods based on the tree ring width and trunk density; A model module is used to obtain three-dimensional point cloud data of multiple trees in the riparian forest, obtain the volumes of the multiple trees based on the three-dimensional point cloud data, and establish a trunk volume model of a single tree; wherein the trunk volume model of a single tree represents the volume morphology of multiple trees in the riparian forest; The ratio module is used to obtain the trunk biomass of individual trees based on the trunk volume model and trunk density of individual trees; calculate the ratio between the trunk volume and branch volume of multiple trees in the riparian forest at different time periods, and obtain the ratio between the trunk volume and branch volume of individual trees at different time periods; obtain the ratio between the trunk biomass and branch biomass of individual trees at different time periods based on the ratio between the trunk volume and branch volume of individual trees at different time periods and the trunk biomass of individual trees; The prediction module is used to obtain the biomass of individual trees based on the radial growth of trees in different time periods and the ratio of trunk biomass to branch biomass of individual trees in different time periods, and combine it with the number of trees counted in the riparian forest to obtain the estimated result of the riparian forest biomass.

[0013] An embodiment of the present invention further provides an electronic device, including a memory and a processor; The memory is used to store computer programs; The processor is used to implement the steps of the above-mentioned method for estimating the biomass of desert riparian forests when executing the computer program stored in the memory.

[0014] An embodiment of the present invention further provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for estimating the biomass of a desert riparian forest.

[0015] The embodiments of the present invention provide a method, device, equipment, and medium for estimating the biomass of desert riparian forests. Compared with the prior art, the methods and devices have the following beneficial effects: The present invention first analyzes the tree ring information and density information at different trunk heights in the riparian forest, and uses radial growth to obtain the annual change in trunk biomass; then, based on the three-dimensional scanning information of the riparian forest trunks, a trunk layered volume algorithm is constructed to calculate the volume of the trunks, and a riparian forest trunk volume model is simultaneously established to calculate the trunk biomass; finally, based on field survey statistical analysis, trunk and branch combination information is obtained, and the trunk and branch biomass distribution pattern is analyzed based on the existing crown biomass model, trunk volume model and trunk density; this process overcomes the difficulty in estimating trunk biomass caused by the large individual differences in Populus euphratica tree shape and the irregular distribution of tree shape by introducing the trunk morphological characteristics and growth characteristics represented by the tree ring information and three-dimensional scanning information of the riparian forest into the estimation of trunk volume, and at the same time, combined with the trunk and branch biomass distribution law of the riparian forest, further considers the individual differences in Populus euphratica tree shape, so as to improve the statistical accuracy of Populus euphratica biomass over a large area as a whole, and provide technical support for forest management and future sustainability research. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the overall process of a method for estimating the biomass of a desert riparian forest provided by an embodiment of the present invention; Figure 2 A schematic diagram of a process for estimating annual trunk biomass of a desert riparian forest biomass estimation method provided by an embodiment of the present invention; Figure 3 A schematic diagram of density sampling at different tree trunk heights for a method for estimating biomass of a desert riparian forest provided by an embodiment of the present invention; Figure 4 A schematic diagram of trunk volume calculation for a method for estimating desert riparian forest biomass provided by an embodiment of the present invention; Figure 5 A schematic diagram of the distribution of the ratio of trunk height to tree height for a method for estimating biomass of a desert riparian forest provided by an embodiment of the present invention; Figure 6 A schematic diagram of a Populus euphratica trunk photograph and three-dimensional point cloud data visualization method for estimating the biomass of a desert riparian forest provided by an embodiment of the present invention; Figure 7 Schematic diagram of the distribution of trunk heights for a method for estimating desert riparian forest biomass provided by an embodiment of the present invention; (a) is a probability density distribution diagram of trunk height differences; (b) is a scatter plot of trunk heights measured using a point cloud and field measurements after removing outliers; Figure 8 Schematic diagram of the volume distribution of Populus euphratica according to a method for estimating the biomass of desert riparian forests provided by an embodiment of the present invention; (a) is a volume distribution diagram of Populus euphratica with different diameters at breast height ranges, and (b) is a scatter plot of the volume of Populus euphratica and the model prediction results; Figure 9 A schematic diagram of a Populus euphratica volume prediction model for a method for estimating desert riparian forest biomass provided by an embodiment of the present invention; Figure 10 A schematic diagram of trunk biomass distribution for a method for estimating biomass of a desert riparian forest provided by an embodiment of the present invention; Figure 11 A schematic diagram of the trunk biomass of Populus euphratica in four sample plots for a method for estimating the biomass of a desert riparian forest provided by an embodiment of the present invention; Figure 12 A schematic diagram of the trunk and branch structure of a method for estimating the biomass of a desert riparian forest provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings. The following description sets forth many specific details to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art can make similar modifications without violating the scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0018] See also Figure 1 The embodiment of the present invention provides a method for estimating the biomass of a desert riparian forest, comprising: 1. Establishment of annual trunk biomass estimation model for different trunk heights.

[0019] 1. Research ideas.

[0020] The study found that the radial growth of the tree rings at different heights of the same Populus euphratica tree was significantly correlated (the average Pearson correlation coefficient was 0.878). Therefore, using the Populus euphratica tree ring information at a certain height can reflect the changing trend of the radial growth of the entire trunk. At the same time, the distribution of Populus euphratica trunk density varies slightly at different parts of the trunk, with an average trunk density of 471.4 kg·m -3 To estimate the annual biomass of Populus euphratica at different heights, we can use the tree ring method to analyze the annual changes in trunk volume and biomass at different heights using the radial growth and density information of different parts of the trunk.

[0021] The new biomass of Populus euphratica trunks is divided into two parts: volume and density. Using tree ring samples and trunk density samples collected in the field, the radial growth and density information at different heights of the trunk are extracted to calculate the trunk biomass. The specific process is as follows: Figure 2 shown.

[0022] 2. Trunk volume sampling.

[0023] Five mature Populus euphratica trees with good growth were selected in the area, and core samples were collected at different heights of the trunk (0.5 m, 1 m, 1.3 m, 1.5 m, and 2 m) using growth cones. A total of 40 cores were sampled from the five trees. In addition, the present invention selected five trees for density sample collection. At different heights of the trunk (0-0.5 m, 0.5-1 m, 1-1.5 m, 1.5-2 m, and 2-2.5 m), part of the bark was removed to obtain tree blocks, and the fresh weight was weighed, resulting in a total of 59 small square samples, such as Figure 3 shown.

[0024] After preliminary dating, radial growth measurements were taken using a tree-ring analyzer (LINTAB) with an accuracy of 0.001 mm. Comparison and correction were performed using a narrow annual skeleton diagram to complete the processing of the tree core samples. After removing tree ring samples that were difficult to identify, 32 core samples were retained, including those at heights of 0.5 m, 1 m, 1.3 m, 1.5 m, and 2 m. Pearson correlation analysis was then performed on the radial growth data obtained at different trunk locations to analyze the correlation between radial growth at different heights within the same tree and the correlation between the width of the tree rings in other Populus euphratica trunks.

[0025] 3. Processing of trunk density samples.

[0026] The method comprises weighing the fresh weight of trunk wood blocks, baking them in an oven at a constant temperature of 85°C to a constant weight, weighing the dry matter of the samples, and then measuring the volume of the trunk density samples using the water displacement method. The density of each wood block is calculated based on the volume and dry weight data of the trunk density samples. The density is then analyzed for each tree and each trunk height to obtain the average density and standard deviation of Populus euphratica at different heights and the average density and standard deviation of all samples.

[0027] 4. Calculation of trunk volume and biomass.

[0028] Taking into account the operability of sampling and the convenience of calculation, the present invention selects tree trunks 0.5 to 3 m above the ground for research. This section of samples is used to study the variation pattern of radial growth and density distribution of tree trunks at different heights, which serves as the basis for studying the biomass of the entire tree trunk. The specific calculation process of the trunk volume is as follows: first, the density of the trunk dry matter is calculated using the sampled density samples; then, the width of the tree rings is extracted, and the volume of the corresponding height section of the Populus euphratica is calculated using the frustum formula. The difference between the volume in year i and the volume in year i-1 ( ), which is the trunk volume occupied by the annual added biomass.

[0029] The radius of the outermost circle of the trunk is used as the radius r (or R), which includes the width of the tree ring from the sampling year before. Using the tree ring sample, the annual radial growth d can be measured, and the trunk radius r corresponding to each year is i , that is, the trunk radius in the previous year is r i-1 =r i -d; then use the volume calculation formula to get v i and v i-1 , the difference between the two is the volume occupied by the newly added biomass of the trunk in year i; Figure 4 As shown, for example: r 1999 、 r 2000 is the trunk radius of the tree at a height of 2 m in 2000 and 1999; R 1999 、R 2000 are the trunk radii at 1.5 m in 2000 and 1999, respectively. The volume occupied by the newly added biomass in the 1.5-2 m section of the trunk height in 2000 can be obtained using the following formula.

[0030] .

[0031] Where: v is the volume of new trunk biomass in year i (m 3 ); h is the trunk length (m); r i 、r i-1 、R i 、R i-1 are the top and bottom radii of the trunk cross section in year i and i-1, respectively (m).

[0032] The calculation formula for annual trunk biomass is: .

[0033] Where: w is the annual increase in trunk biomass (kg); is the trunk density (kg·m-3); d and D are the additional radial growth of the trunk in year i (m).

[0034] 2. Optimization of trunk biomass estimation model.

[0035] 1. Research ideas.

[0036] This paper uses a 3D scanning device (Google Tango mobile phone) to obtain 3D information of tree trunks, achieving non-destructive sampling. It also proposes an algorithm that can directly calculate the volume of point cloud data without 3D reconstruction of the point cloud data. A trunk volume model was established using the point cloud information of 91 Populus euphratica tree trunks. The present invention uses a LenovoPhab 2 Pro mobile device equipped with Google Tango for 3D scanning.

[0037] 2. Field sampling.

[0038] Field data sampling was carried out in the area to measure the trunk height and tree height information of a total of 100 Populus euphratica trees, including 91 scanned Populus euphratica trees.

[0039] ① Collection of tree trunk three-dimensional point cloud information.

[0040] The desert riverbank forest section was divided into several sections. Approximately 30 Populus euphratica trees were selected from each section for scanning, generating 3D point cloud data. Information such as trunk diameter at breast height, tree height, and base diameter was measured. Using the Android 3D Scanning app (https: / / matterport.com / ), powered by Google Tango and boasting a scanning accuracy of 2.5 cm (according to the app's official documentation), 360-degree scans of individual Populus euphratica trunks were performed to generate 3D point cloud datasets. The scanning position and viewing angle could be adjusted during the scanning process. On average, over 10,000 point clouds were collected for each tree trunk, encompassing a wealth of tree shape information, as shown in Table 1.

[0041] Table 1 DBH range and average number of point clouds of Populus euphratica sampled from point cloud data According to the tree height and trunk height information of all Populus euphratica, the ratio of trunk height to total tree height is 0.417. Figure 5 shown.

[0042] ② Sampling of Populus euphratica quadrat at Yingsu section.

[0043] Four 50×50 m plots were set up near the monitoring wells of the Yingsu section in an area with a large number of Populus euphratica trees. They were F1 (42°26′51.5″N, 87°56′17.0″E), F6 (40°24′38″N, 87°56′17.5″E), F8 (40°24′10.4″N, 87°56′15.2″E), and F11 (40°24′0.17″N, 87°56′28.8″E). The DBH, basal diameter, tree height, and crown width of all Populus euphratica trees in the plots were measured, as shown in Table 2.

[0044] Table 2 Basic ecological indicators of Populus euphratica in the sample plot 3. Data processing and calculation.

[0045] This paper designs an algorithm for calculating tree trunk volume using 3D point cloud data. The main idea of this algorithm is to vertically divide the tree trunk point cloud dataset into n layers, calculate the volume of each layer separately, and then superimpose the volumes of the n layers to obtain the volume of the entire tree trunk. The specific steps of the point cloud data algorithm are as follows: In the first step, in the volume algorithm, a point cloud dataset of a tree is divided into n slices in the vertical direction (100 layers in the calculation process of the present invention).

[0046] In the second step, in each slice, two upper and lower polygons are generated to determine the periphery of the upper and lower planes of the polygon.

[0047] The third step is to find the corresponding points with the top and bottom in the vertical direction.

[0048] The fourth step is to generate an area function of the polygonal area c(r) .

[0049] The fifth step is to construct an area integral function based on the area function to calculate the slice volume.

[0050] In the sixth step, the volumes of 100 slices are calculated and superimposed.

[0051] In the seventh step, the algorithm of single trunk volume is looped until the datasets of all trunks are calculated.

[0052] In this algorithm, the volume of irregularly shaped trunks can also be calculated together with other Populus euphratica. Dividing the trunk dataset into multiple layers can ensure that the irregular parts of the trunk are included in the volume results.

[0053] 4. Verification of volume model and measurement accuracy.

[0054] ①Model construction.

[0055] The present invention uses linear regression analysis to construct a volume model and also uses the Adaboosting model algorithm in machine learning for regression modeling. The coefficient of determination (R 2 ), also known as the fit index, is calculated by the total sum of squares (TSS) and the residual sum of squares (RSS), according to R 2 The applicability of the established model is judged by the significance test of the size and equation, R 2 The larger the value, the more reliable the model is.

[0056] ②Verification of measurement accuracy.

[0057] Since Populus euphratica is a protected tree species in the Tarim River, it is impossible to measure the trunk volume or biomass to verify the accuracy of this research model; therefore, the present invention analyzes the error between the trunk height obtained by the 3D scanner and the trunk height measured manually to verify the accuracy of the collected samples of the 3D scanning.

[0058] 3. Implementation of tree trunk point cloud data processing algorithm.

[0059] 1. Research ideas.

[0060] A scanning device was used to obtain a 3D point cloud of the tree trunk, which was then visualized using MeshLab. The point cloud data was matched with images of the tree. Excess point clouds were cleaned up to retain only the trunk portion, and the trunk volume algorithm was used to calculate the volume. Finally, the various parameters of the Populus euphratica were used to fit the model with the volume to obtain a reliable model with reasonable accuracy. Based on the 3D data, the fitted model can include the morphological information of the trunk, reflect the individual differences of the Populus euphratica, and have a high model accuracy.

[0061] 2. 3D point cloud data visualization.

[0062] ① Visualization of tree trunk 3D point cloud data.

[0063] This paper uses MeshLab software to visualize the tree trunk point cloud data. Figure 6 It can be seen that the three-dimensional point cloud information has a high degree of restoration of the real tree trunk. Even if the trunk is bent or the tree shape is irregular, the poplar trunk information can be effectively collected. Irregular trunk shapes are more common in the poplar forests in the lower reaches of the Tarim River. Using a three-dimensional scanner for sampling, information such as the bending and bifurcation of the trunk can be included.

[0064] ②Verification of scanner sampling accuracy.

[0065] like Figure 7 As shown in Figure 3, from the probability density distribution of the trunk height difference, the overall difference presents a normal distribution.

[0066] 3. Tree trunk volume model based on three-dimensional model.

[0067] ①Linear regression.

[0068] The present invention combines Populus euphratica tree height, trunk height, trunk base diameter and breast diameter with volume, uses linear regression analysis in Scikit-learn to fit the model, uses 2 / 3 of the data to train the model, and uses 1 / 3 of the data set to verify the model results; Figure 8 The red data points are the model prediction results, which have the same trend as the training data. The volume model equation of Populus euphratica trunk is obtained through model fitting and verification. The linear regression model equation of Populus euphratica volume is: .

[0069] Where: v (m 3 ) is the volume; DBH (cm) is the diameter at breast height of Populus euphratica; BD (cm) is the base diameter of the trunk; H (m) is the tree height; SH (m) is the trunk height; DH 2 is a derived factor of tree height (m) and diameter at breast height (m).

[0070] like Figure 9 As shown, the linear model prediction value (P) and the Tango scan volume (T), P is the volume predicted by the linear equation, and T is the trunk volume calculated by the scanner measurement. It can be seen that the model has high accuracy.

[0071] ②Ada boosting model.

[0072] The present invention also uses Ada boosting to fit the data model, using 80% of the data set for training the model and 20% for testing the model; after model testing, it was found that R 2 The accuracy of the test model is affected by the number of fitted parameters. When using H, DBH and BH2, R 2 The value is 0.715; using DBH and BH2 for gradient regression, the model prediction accuracy R 2 is 0.740, MSE is 0.0064, and the model accuracy R 2 The values are slightly lower than those of the simple linear model.

[0073] 4. Populus euphratica trunk biomass results.

[0074] ①Total trunk biomass based on 3D scanning results.

[0075] The volume of 91 Populus euphratica trees was calculated using the point cloud data algorithm. Combined with the density (471.4 kg·m-3), the trunk biomass was calculated. The average trunk biomass of these Populus euphratica trees was 103.7 kg, and the trunk biomass of Populus euphratica trees was concentrated between 87.47 and 119.96 kg. Figure 10 shown.

[0076] ② Biomass of Populus euphratica trunks in four monitoring wells of the Yingsu section.

[0077] The biomass of Populus euphratica from four plots in the Yingsu section was analyzed. The average trunk biomass of a single tree was 136.84 kg. The trunk biomass results calculated based on the two samplings were compared. It was found that the overall trend of the trunk biomass distribution of Populus euphratica was relatively consistent, showing a positive skewed distribution. Figure 10 and Figure 11 As shown in the figure, there are a total of 11 monitoring wells, F1 to F11, which are arranged in the vertical direction of the river channel. Among them, due to the bifurcation of the river channel at the Yingsu section, F6 is the monitoring well farthest from the two river channels, and the closer it is to the river channels, the closer it is to the river channels. According to the four sampling plots, at different distances from the river, the number, growth and total trunk biomass of Populus euphratica in the same area of the plots are different. The total trunk biomass and average biomass of F1 and F11 are higher than those of the other two plots.

[0078] 4. Rapid estimation of biomass based on the typical Populus euphratica branch axis model.

[0079] 1. Research ideas.

[0080] The present invention compares the biomass distribution of typical Populus euphratica constructed based on branch axis statistics and the crown-trunk biomass ratio in four sample plots in England and Russia. The trunk biomass of the typical Populus euphratica accounts for 78% of the sum of the trunk-branch biomass, and the ratio of the trunk biomass to the aboveground biomass in the four sample plots in England and Russia is all above 83%.

[0081] The branch-axis composition of Populus euphratica trees reflects their growth morphology and condition, and reflects the biomass of branches and crowns, as well as their distribution patterns. Using a large amount of branch-axis statistical data, a branch-axis composition model for a typical Populus euphratica tree is constructed to estimate biomass distribution patterns. For desert Populus euphratica forests, where growth characteristics vary widely, a large number of statistical experiments using non-destructive sampling can quickly and easily estimate the total biomass of a given area.

[0082] 2. Data and research methods.

[0083] The field sampling was divided into two parts: the first was to count the number and distribution of branches of various levels of a large number of Populus euphratica at Yingsu section and Kumtuge in March 2017, and count the number and base diameter range of primary branches (branches growing on the main trunk and before the bifurcation of primary branches), secondary branches (branches growing on primary branches and before the bifurcation of secondary branches), and tertiary branches (branches growing on secondary branches); the Populus euphratica was counted in four large sample plots of Populus euphratica established in Yingsu; the second was to select 30 Populus euphratica trees at Yingsu section in August 2017, and measure the base diameter, diameter at breast height, diameter at the bifurcation of the trunk, and the number, base diameter and branch length of branches of various levels; the two samplings included the diameter at breast height and crown width of 355 Populus euphratica trunks; the branch length, base diameter and number of primary and secondary branches of 150 Populus euphratica trees; and the number and base diameter of tertiary branches of 79 Populus euphratica trees.

[0084] ①Construct the trunk-branch composition of a typical Populus euphratica based on statistical data.

[0085] Using the trunk diameter at breast height, tree height and crown width information of 355 Populus euphratica trees collected in the field; the branch length, base diameter and number of primary and secondary branches of 150 Populus euphratica trees; the number and base diameter of tertiary branches of 79 Populus euphratica trees, the basic conditions of Populus euphratica trunks and branches were analyzed, and a combined trunk-branch axis model of a typical Populus euphratica was constructed as the basis for studying trunk-branch biomass, such as Figure 12 Shown is a diagram of the trunk and branch structure.

[0086] ②Study on biomass allocation.

[0087] Based on the information collected from the four large Populus euphratica plots in England and the Soviet Union, trunk biomass was calculated using the Populus euphratica volume model proposed above combined with trunk density. The aboveground biomass of Populus euphratica is divided into two parts: crown biomass and trunk biomass. Trunk biomass = trunk volume × density. The volume is calculated using the volume model, that is, the following formula.

[0088] M = 0.016A2+2.291A+11.084 .

[0089] Where: Mcrown is the crown biomass (kg); A is the crown area (m 2 ), A=πr² = , where a and b are the crown widths in the east-west and north-south directions (m), respectively.

[0090] 3. Statistics of the number and length of Populus euphratica branches.

[0091] ① Statistics of the number and length of Populus euphratica branches.

[0092] Through field statistics, a total of 335 Populus euphratica trees were collected, including trunk diameter at breast height and tree height, and base diameter, branch length, and number of primary and secondary branches of 157 trees. Among them, 79 Populus euphratica trees had base diameter and number of tertiary branches, and 30 trees had branch length of tertiary branches. The trunk diameter at breast height of Populus euphratica ranged from 20 to 130 cm, and the tree height ranged from 3 to 11 cm. m; Statistics show that the morphology of Populus euphratica is closely related to the combined morphology of the main trunk and primary branches, and can be roughly divided into two categories: (1) with one main trunk, the number of primary branches is greater than 2, and the base diameter of the secondary branches decreases gradually with the increase of the number of branches; (2) with one main trunk, but the main trunk is divided into two at a lower trunk height (even bifurcated at a position close to the base diameter), and the two branches are regarded as primary branches. The base diameter of each primary branch is close to the breast diameter of an adult Populus euphratica, and the base diameters of other secondary branches are also relatively large; Analysis of all statistical data shows that the number of primary branches of Populus euphratica is mostly between 2 and 5, with an average of 3.8 branches, and the base diameter is between 5 and 25 cm; the number of secondary branches is between 10 and 50, with an average of 22 branches, and the base diameter of secondary branches is between 2 and 6 cm; the number of tertiary branches is between 30 and 120, with an average of 87 branches; the average base diameter is 1.5 cm; the number of primary branches is concentrated in 3~3.5 branches, the number of secondary branches is concentrated in 11~30 branches, and the number of tertiary branches is concentrated in 61~80 branches; the difference in the number of primary and secondary branches is related to the age of the tree and the size of the base diameter. Young trees in the growth stage usually have more primary and secondary branches and the base diameters of primary and secondary branches are smaller; the number of tertiary branches is large and difficult to count, but the density of tertiary branches can well reflect the growth condition of Populus euphratica. Populus euphratica with dense tertiary branches often has a more rounded crown shape and grows better, and the denser the crown; the sparser the tertiary branches, the more incomplete the crown shape and the worse the growth.

[0093] ② Statistics of the base diameter and branch length of the third-level branches of 30 Populus euphratica trees.

[0094] The present invention uses the collected information on the base diameter and branch length of the primary, secondary and tertiary branches of 30 Populus euphratica plants. The analysis shows that the base diameter of these 30 Populus euphratica plants ranges from 4 to 22 cm; the base diameter of the secondary branches is mainly concentrated in the range of 4 to 13 cm, with a few exceeding 15 cm; the base diameter of the tertiary branches is mainly concentrated in the range of 1 to 4 cm; The length of primary branches of Populus euphratica ranges from 0.831 to 1.189 m, the length of secondary branches ranges from 0.651 to 0.0749 m, and the length of tertiary branches ranges from 1.25 to 1.277 m. Using the cylinder volume calculation formula to estimate branch volume can quickly and easily understand branch biomass.

[0095] 4. Biomass and biomass distribution ratio of typical Populus euphratica trunks and branches.

[0096] Using statistical data on Populus euphratica branches and axes, a typical Populus euphratica trunk-branch-axis model was established. The volume of branches was estimated with the help of the cylinder volume algorithm, and the biomass and distribution ratio were further calculated. The crown height of the whole tree was set as h (m), the lengths of the first, second, and third-level branches were set as h1, h2, and h3, respectively, the volume of the trunk was set as v1, and the volumes of the first, second, and third-level branches were set as b1, b2, and b3, respectively. The average of each attribute in the statistics was used as the calculated value. The branch lengths of the three levels were 1.01 m, 0.7 m, and 1.25 m, respectively, as shown in Table 3.

[0097] Table 3 Estimated biomass of typical Populus euphratica branches and trunks Calculations showed that the trunk and branch biomass of a typical Populus euphratica were 174.57 kg, 20.46 kg, 12.5 kg, and 6.29 kg, respectively, totaling 213.87 kg. The branch biomass accounted for 18.38% of the trunk biomass, and the trunk biomass accounted for 81.62% of the trunk-branch biomass. The first-, second-, and third-order branches accounted for 9.57%, 5.84%, and 2.94% of the total trunk-branch biomass, respectively. The ratio of the trunk to the first-, second-, and third-order branches was 27.75:3.25:1.99:1.

[0098] 5. Model-based crown-to-trunk biomass ratio.

[0099] This study used crown biomass models, trunk volume models, and trunk density to calculate crown and trunk biomass in four large plots from England and the Soviet Union, deriving the crown-to-trunk biomass ratio for Populus euphratica. Crown biomass varied across plots. The total crown biomass within the F11 plot reached 2374.27 kg, with an average trunk biomass of 26.02 kg per tree. The total crown biomass within the F1 plot reached 1264.59 kg, with the highest average biomass at 32.02 kg. The average and total biomass within the F6 and F8 plots were relatively close. Overall, the crown biomass of Populus euphratica trees within the four plots ranged primarily between 20 and 25 kg, with the F6 plot showing slightly lower crown biomass distribution than in the other plots.

[0100] The average proportion of trunk biomass to aboveground biomass in the four plots (F1, F6, F8, and F11) was 84%, 90%, 89%, and 83%, respectively. The crown-to-trunk biomass ratios for F11 and F1 were 0.3518 and 0.1759, respectively, which were higher than the crown ratios in the other two plots. The distribution ratio of trunk and crown biomass of Populus euphratica varied across different site conditions. Aboveground biomass in desert vegetation varied with distance from a river, with crown biomass accounting for a larger proportion in plots closer to a river than in plots farther from a river. Plots farther from a river had a higher proportion of trunk biomass to total biomass. The trunk-to-crown biomass ratio may also be related to the growth stage and age of the Populus euphratica.

[0101] The present invention can incorporate the growth morphology of Populus euphratica trunks into the volume estimation, thereby improving the accuracy of Populus euphratica trunk volume estimation; combining trunk-branch axis statistical information, the distribution pattern of Populus euphratica biomass and Populus euphratica branch-trunk biomass is obtained; rapid estimation of Populus euphratica biomass based on a typical Populus euphratica model can facilitate rough estimation of Populus euphratica aboveground biomass in the field, and also provide technical support for biomass estimation of other rare and protected tree species. The volume calculated by the present invention using three-dimensional trunk information can include morphological characteristics, achieving more accurate trunk biomass estimation; the research method used by the present invention to calculate trunk volume based on a three-dimensional mobile phone scanner can overcome the problem of difficulty in biomass estimation caused by large differences in Populus euphratica tree shapes, and is non-destructive, and can be applied to biomass research of other rare tree species; at the same time, it enriches the research method of applying three-dimensional data to forestry, has good promotion and application value, and provides technical support for forest management and future sustainability research.

[0102] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the patent for this invention shall be determined by the appended claims.

Claims

1. A method for estimating the biomass of desert riparian forests, characterized in that: The following steps are involved: Obtain the tree ring width and trunk density at different heights of riparian forest trees, and based on the tree ring width and trunk density, obtain the radial growth of trees in different time periods; Obtaining 3D point cloud data of multiple trees in a riparian forest, obtaining the volumes of the multiple trees based on the 3D point cloud data, and establishing a trunk volume model for each individual tree; wherein the trunk volume model for each individual tree characterizes the volumetric morphology of the multiple trees in the riparian forest; The trunk biomass of individual trees was obtained based on the trunk volume model and trunk density of individual trees. The ratio between the trunk volume and branch volume of multiple trees in riparian forests at different time periods was calculated, and the ratio between the trunk volume and branch volume of individual trees at different time periods was obtained. Based on the ratio between the trunk volume and branch volume of individual trees at different time periods and the trunk biomass of individual trees, the ratio between the trunk biomass and branch biomass of individual trees at different time periods was obtained. The biomass of individual trees was obtained based on the radial growth of trees in different time periods and the ratio of trunk biomass to branch biomass of individual trees in different time periods. Combined with the number of trees counted in the riparian forest, the estimated result of the riparian forest biomass was obtained.

2. The method for estimating the biomass of a desert riparian forest according to claim 1, characterized in that: The acquisition of the radial growth amount of the tree in different time periods includes: Using the tree ring samples and trunk density samples collected from the riparian forest in the field, the density of trunk dry matter was obtained using the collected density samples; Extract the tree ring width and use the frustum formula to calculate the volume of the corresponding height segment of Populus euphratica. The difference between the volume in year i and the volume in year i-1 is: , the volume difference is the radial growth of the tree in year i.

3. The method for estimating the biomass of a desert riparian forest according to claim 1, characterized in that: The obtaining of volumes of the plurality of trees based on the three-dimensional point cloud data includes: Obtain 3D point cloud data of multiple riparian forest tree trunks, and segment the 3D point cloud data of the tree trunks into n layers in the vertical direction; The 3D point cloud data of a tree is divided into n slices in the vertical direction. In each slice, two upper and lower polygons are generated to determine the perimeter of the upper and lower planes of the polygon, and the corresponding points with the top and bottom surfaces are found in the vertical direction. Based on the periphery of the upper and lower planes of the determined polygon, as well as the corresponding points on the top and bottom surfaces, an area function of the polygonal area is generated. Based on the area function, an area integral function is constructed to calculate the volume of the slice. The volumes of n slices are obtained and superimposed to obtain the volume of the entire tree trunk.

4. The method for estimating the biomass of a desert riparian forest according to claim 3, characterized in that: The area integral function is: ； Where: v represents the volume of the entire tree trunk; r represents the trunk radius of the smallest periphery of the slice; r(x) represents the trunk radius of the largest periphery of the slice; c(r i ) represents the i-th polygonal area.

5. The method for estimating the biomass of desert riparian forest according to claim 4, characterized in that: The construction of the trunk volume model of the single tree includes: Based on the 3D point cloud data, the diameter at breast height, base diameter, tree height and crown width of each riparian forest tree were obtained; Based on the trunk volume of each riparian forest tree and the corresponding diameter at breast height, base diameter, tree height, and crown width information of each riparian forest tree, linear regression analysis in Scikit-learn software was used to fit the model and obtain the trunk volume model; the trunk volume model is: ； Where: v is the volume; DBH is the diameter at breast height; BD is the base diameter of the trunk; H is the height of the tree; SH is the height of the trunk; DH 2 Derivative factors representing tree height and diameter at breast height.

6. The method for estimating the biomass of a desert riparian forest according to claim 1, characterized in that: The method of obtaining the ratio between the trunk biomass and the branch biomass of a single tree in different time periods includes: Field statistics were collected on the diameter at breast height, base diameter, tree height, and crown width of multiple riparian forest tree trunks. One part collected information on the length, base diameter, and number of primary and secondary branches in the riparian forests, while another part collected information on the number and base diameter of tertiary branches in the riparian forests. Furthermore, information on the trunk and branch combinations of multiple riparian forest trees was obtained. In statistics, the difference in the number of primary and secondary branches is related to the tree age and base diameter. The base diameter of primary and secondary branches is smaller, and the number of tertiary branches is larger. Trees with dense tertiary branches have a rounded crown shape, good growth, and dense crowns. Trees with sparse tertiary branches have an incomplete crown shape and poor growth. Based on the statistical number of primary, secondary and tertiary branches, and using the crown biomass model, trunk volume model and trunk density, the trunk and branch biomass distribution ratios and the trunk and crown biomass distribution ratios of individual trees in riparian forests in different time periods were obtained respectively; based on the trunk and branch biomass distribution ratios and the trunk and crown biomass distribution ratios, the ratios between trunk biomass and branch biomass of individual trees in different time periods were obtained.

7. A device for estimating the biomass of desert riparian forests, characterized in that: include: The tree ring module is used to obtain the tree ring width and trunk density of riparian forest trees at different heights, and to obtain the radial growth of trees in different time periods based on the tree ring width and trunk density; A model module is used to obtain three-dimensional point cloud data of multiple trees in the riparian forest, obtain the volumes of the multiple trees based on the three-dimensional point cloud data, and establish a trunk volume model of a single tree; wherein the trunk volume model of a single tree represents the volume morphology of multiple trees in the riparian forest; The ratio module is used to obtain the trunk biomass of individual trees based on the trunk volume model and trunk density of individual trees; calculate the ratio between the trunk volume and branch volume of multiple trees in the riparian forest at different time periods, and obtain the ratio between the trunk volume and branch volume of individual trees at different time periods; obtain the ratio between the trunk biomass and branch biomass of individual trees at different time periods based on the ratio between the trunk volume and branch volume of individual trees at different time periods and the trunk biomass of individual trees; The prediction module is used to obtain the biomass of individual trees based on the radial growth of trees in different time periods and the ratio of trunk biomass to branch biomass of individual trees in different time periods, and combine it with the number of trees counted in the riparian forest to obtain the estimated result of the riparian forest biomass.

8. An electronic device, characterized in that: include: memory and processor; The memory is used to store computer programs; The processor is configured to implement the steps of a method for estimating the biomass of a desert riparian forest as claimed in any one of claims 1 to 6 when executing the computer program stored in the memory.

9. A computer-readable storage medium, characterized in that Used to store a computer program, which, when executed by a processor, implements the steps of a method for estimating the biomass of a desert riparian forest as claimed in any one of claims 1 to 6.

Images