Search

CN-115239512-B - Method for determining age of single wood in natural forest stand

CN115239512BCN 115239512 BCN115239512 BCN 115239512BCN-115239512-B

Abstract

The invention discloses a method for determining the age of single wood in a natural forest stand, which comprises the steps of selecting reserved wood with multi-period actually measured breast diameters in the natural forest stand, grading the forest at the annual growth rate of the breast diameters, constructing a forest grade dumb variable based on a theoretical growth equation, adopting a starting point constraint, establishing a statistical model of the multi-period single wood breast diameters and corresponding investigation time intervals, and reversely pushing the age of the single wood through the model. According to the invention, the problem of difficulty in obtaining the age of single wood in the natural forest stand is solved by reserving multi-period breast diameter data of the wood and combining the differentiation condition of the wood in the forest stand under the condition of not influencing the growth of the wood and reducing investigation work, and age data is provided for building a natural forest reserve growth model, so that the method is beneficial to monitoring the dynamic change of natural forest resources and evaluating the stand quality of the forest stand.

Inventors

  • LI HAIKUI
  • HUANG JINJIN
  • ZHANG YIRU

Assignees

  • 中国林业科学研究院资源信息研究所

Dates

Publication Date
20260508
Application Date
20220725

Claims (7)

  1. 1. Selecting reserved wood with multi-period actually measured breast diameters in a natural forest stand, grading the forest at the annual growth rate of the breast diameters, constructing a forest grade dumb variable based on a theoretical growth equation, adopting a starting point constraint, establishing a statistical model of the multi-period breast diameters of the single wood and corresponding investigation time intervals, and reversely pushing the age of the single wood through the model; The method for determining the age of the single wood in the natural forest stand comprises the following steps: step one, fitting the relation between the breast diameters of the multi-period single wood and the corresponding investigation time intervals, and determining model parameters: based on a theoretical growth equation, a relation model of multi-stage single wood breast diameters and corresponding investigation time intervals is established, wherein the total number of the single wood breast diameter growth models is 2, and the form is as follows: The first model is based on MITSCHERLICH growth equation, adopts starting point constraint, and reflects tree classification on the parameter A, and the model is as follows: ), And a second model, namely taking a Richards growth equation as a basic model, adopting a starting point constraint, and simultaneously embodying tree classification on the parameters A and C, wherein the model is as follows: ), Wherein, the , , Wherein: The actual measurement chest diameter of the ith plant in the jth period is reserved, m is the investigation period number, The actual measurement chest diameter of the ith plant in the initial stage of wood investigation is reserved, For the time interval between the j-th phase and the initial phase of investigation, 、b、 As a parameter of the model, it is possible to provide, 、 、 As a dummy variable for the final tree grading, when the retained wood is the dominant wood in the natural stand, , When the retained wood is an average wood in a natural stand, , When the reserved wood is pressed wood in the natural forest stand, , , Is the difference of the adjacent forest grade level parameter a, Is the difference value of the parameter c between adjacent tree grades; model parameters are obtained after fitting a single wood chest diameter growth model: Step 1-1, selecting reserved wood with multi-period actually measured breast diameters by taking dominant tree species in a certain natural forest sample plot as a research object, and taking breast diameter data of the reserved wood as modeling data, wherein each reserved wood has a unique sample wood number; Step 1-2, based on a theoretical growth equation, establishing a relation between the breast diameters of the multi-stage single wood and corresponding investigation time intervals; Step 1-3, adopting a starting point constraint basic chest diameter growth model, and converting the model form; Step 1-4, calculating annual growth rate of retained wood breast diameters according to breast diameters in the first and last two stages, and carrying out initial tree grading on the retained wood in the sample plot by adopting ordered sample clustering; Step 1-5, combining an initial tree grading result, and taking a Richards growth equation as a grading equation to obtain a final tree grading; Step 1-6, constructing a dummy variable in a grading way by using a final forest, parameterizing model parameters, and determining a final single-wood breast diameter growth model; step 1-7, fitting by adopting a least square method to obtain model parameters; Step two, reversely pushing a single wood age calculation formula, and estimating the single wood age by using model parameters and first-period chest diameter data: step 2-1, reversely pushing an age calculation formula by a single wood chest diameter growth model; step 2-2, determining the tree grade of the single tree in the natural forest, and corresponding to the dummy variable parameter value; step 2-3, calculating the initial age of the single wood according to the initial chest diameter data; and 2-4, obtaining the ages of the single wood in the rest periods according to the time intervals of the corresponding investigation periods.
  2. 2. The method for determining the age of a single wood in a natural forest stand according to claim 1, wherein the modeling data of step 1-1 is selected from the group consisting of: The method comprises the steps of taking dominant tree species in a natural forest sample plot as a study object, taking continuous investigation data of a fixed time interval as basic data, taking a first investigation period as a starting point and a last investigation period as an end point, taking m investigation periods as total, taking the single wood participating in modeling as continuous time sequence data, namely m actually measured breast diameter data in m investigation periods, and taking time sequence data of partial sample wood in the sample plot in m investigation periods as incomplete due to tree competition, namely m actually measured breast diameter data of each reserved wood in the screening plot.
  3. 3. The method for determining the age of single wood in natural forest stands according to claim 1, wherein in the steps 1-2 and 1-3, a basic chest diameter growth model is established and a starting point constraint is adopted, and the method comprises the following steps: Step three, adopting a starting point constraint basic chest diameter growth model: And 3-1, respectively establishing a relation between the two on the basis of MITSCHERLICH growth equation and Richards growth equation by taking the multi-stage breast diameter of the reserved wood as a dependent variable and the corresponding investigation time interval as an independent variable to obtain a basic single-wood breast diameter growth model, wherein the model is constructed as follows: ), ), Wherein: Preserving the initial age of the tree for the ith plant, a being a parameter reflecting the growth potential of the tree, b being a parameter related to the growth rate of the tree species, c being a parameter related to the assimilation power exponent w, ; Step 3-2, restraining the model to pass through the starting point on the basis of the formula (5) and the formula (6) in order to enhance the stability of the model , ) The model is constructed as follows: adopts a starting point constraint type (5): When the model passes through the starting point, ), Formula (5) and formula (7) combine to eliminate unknown age parameters Then, a single wood breast diameter growth model constrained by the starting point is obtained, and the model is as follows: ), similarly, when equation (6) is constrained by the starting point, the model transforms as follows: 。
  4. 4. The method for determining the age of single wood in natural forest stand according to claim 3, wherein in steps 1-4 and 1-5, the degree of differentiation of the forest is different due to resource contention in the natural forest, and in order to reveal the phenomenon, the tree is classified by using the annual growth rate of the breast diameter, 3 classes of trees are replaced by 3 classes of trees, dominant, average and pressed, and the final tree classification result is obtained, which comprises the following steps: step four, determining a final tree grading result: step 4-1, marking each reserved wood with a unique sample wood number, and arranging the reserved wood in an ascending order according to the sample wood number; Step 4-2, calculating annual growth rate v of the breast diameter of each reserved wood by adopting the breast diameters of the first and the last stages; , Wherein: The annual growth rate of the wood breast diameter is reserved for the ith plant, The actual measurement chest diameter of the ith plant in the mth stage is reserved, The mth period is the last period of the first and last investigation periods; Step 4-3, arranging annual growth rates v of all reserved wood breast diameters in the sample plot in a descending order, and grading initial woods by adopting ordered sample clustering to divide the reserved wood in the sample plot into 3 grades; The ordered sample clustering algorithm classifies the samples by adopting an optimal segmentation method, a minimum error function recurrence formula of the ordered sample clustering algorithm is shown as a formula (11), and then the samples are classified according to a minimum error function value; (11) Wherein: A loss function representing the classification is presented, A classification scheme is represented by the formula, For the number of categories to be classified, Which represents the number of samples of a certain type, Indicating the number of plants in the sample plot where the wood remains, Representing the sum of squares of the deviations of a certain class of samples; 4-4, based on Richards growth equation, carrying out parameterization on the parameters a and c by a method of constructing dummy variables at equal intervals on the initially divided 3 forest grades, constructing a relation between multi-period chest diameters of the reserved trees and corresponding investigation time intervals, and obtaining a final tree grading result of each reserved tree in a sample area after fitting, wherein the grading equation is shown as a formula (12), in the final tree grading result, grade 1 represents reserved trees as dominant trees in the sample area, grade 2 represents reserved trees as average trees in the sample area, and grade 3 represents reserved trees as pressed trees in the sample area; ), Wherein, the , , 、 、 For the level dumb variable of the initial tree grading, if the reserved tree is level 1 when the initial tree grading is divided , If the reserved wood is of grade 2 , If the reserved wood is of grade 3 , , Is the difference of the adjacent forest grade level parameter a, Is the difference of the adjacent tree grade level parameter c.
  5. 5. The method for determining the age of single wood in natural forest stand according to claim 4, wherein in the steps 1-6, in order to improve the applicability of the single wood breast diameter growth model, reflecting the influence of the differentiation degree of the forest on the single wood breast diameter growth, introducing the final tree classification result into the formula (8) and the formula (9) by a dummy variable method to obtain the final single wood breast diameter growth model, wherein the model is constructed as follows: Based on the formula (8), the final tree grading result is subjected to a method of constructing dummy variables at equal intervals, and the parameter a is parameterized to obtain a model I: , based on the formula (9), the final tree grading result is subjected to a method of constructing dummy variables at equal intervals, and parameters a and c are parameterized to obtain a model II: , Wherein, the , 。
  6. 6. The method for determining the age of single wood in natural forest stand according to claim 5, wherein in the step 2-1, the derivation process of the calculation formula of the back-push age of the single wood chest diameter growth model is as follows: Based on model one: Is eliminated by the formula (5) and the formula (7) Then, reconstructing a forest grade dummy variable to obtain a model I, and reversely calculating a formula by a formula (5) containing unknown initial age parameters: , in the formula, Reserving the age of the ith plant in the jth stage of wood; Because the parameter a is re-parameterized by adopting a dummy variable method, a is replaced by A, and a final single wood age calculation formula can be obtained: , when the chest diameter is changed from the first stage In the course of the calculation of the time, I.e. the calculation formula becomes as follows: , based on the second model: the formula for calculating the age of the individual wood is as follows:
  7. 7. the method for determining the age of single wood in natural forest stand according to claim 6, wherein in the step 2-2, determining the parameter value of the forest grade dummy variable comprises the following steps: When the initial forest age is estimated for the reserved wood participating in modeling, the value of the parameter A can be determined only by definitely determining which type of forest the reserved wood belongs to according to the final tree grading result because the tree grading is carried out before modeling, if the reserved wood is dominant wood in the sample area, the method can be used for determining the value of the parameter A If the reserved wood is average wood in the sample area, then If the reserved wood is pressed wood in the sample area, then ; For other sample woods which are dominant tree species in the same sample plot but do not meet the complete time sequence data in m investigation periods, namely the sample woods have k actually measured breast diameter data in m investigation periods, 1< k < m; step five, determining the tree grade of the sample wood in the sample land: Step 5-1, assuming that the wood sample is dominant in the sample plot, parameters ; Step 5-2, according to step (14), from the first-stage chest diameter Estimate the initial age of the wood sample as dominant wood ; , Step 5-3, according to (14), from the terminal chest diameter Estimate the end age of the sample wood as dominant wood ; , Step 5-4, calculating the difference value between the calculated first-stage and last-stage forest age interval and the actual age interval when the sample wood is dominant wood; , Wherein, the The first and last two-period investigation time intervals of the sample wood with k actually measured breast diameters in m investigation periods are provided; And 5-5, respectively supposing that the sample wood is average wood and pressed wood in the sample land, repeating the steps, and obtaining the forest grade of the sample wood in the sample land if the OSC value is minimum.

Description

Method for determining age of single wood in natural forest stand Technical Field The invention relates to the technical field of forestry, in particular to a method for determining the age of single wood in natural forest stand. Background The stand structure of natural forest is complex, and has the characteristics of mixing, layering, different ages and the like. The tree species in the forest are various in composition, the ages of the trees are different, and the competition relationship among the trees and the growth change of the trees are continuously changed along with the increase of the ages of the trees. Compared with artificial forests of the same age, the spatial-temporal distribution pattern of each constituent component of the forest group in the natural forests is more reasonable, the anti-interference capability is strong, and the species diversity is richer. The single wood age in the natural forest stand is an important basis for researching the growth rule of the tree, the stand updating succession dynamic state and evaluating the standing quality, but the tree age distribution gradient in the natural forest stand is large, the previous method for determining the single wood age is various, wherein a destructive analysis wood disc measuring method can obtain more accurate single wood age, but the field actual measurement workload is large and the tree is damaged, and an instrument measuring method is high in cost but low in efficiency and cannot be popularized and applied on a large scale, so that the method for predicting the tree age by constructing a model is a feasible simple method. A great number of researches show that a close growth relationship exists between the breast diameter of the forest and the age of the forest, the previous researches adopt a space instead of time method, the breast diameter growth process is simulated by obtaining diameter and age data through trunk analysis, and a tree age checking table is manufactured, but due to the limitation of modeling data, the accuracy of the model is not high and the application range is smaller. The breast diameter data is a common investigation factor in a tree growth model, has the advantages of small measurement error, easy acquisition and the like, can accurately reflect the breast diameter growth and stand growth dynamics of the tree, and lays a foundation for building a related breast diameter growth model. Therefore, the invention constructs the single-tree breast diameter growth model by adopting the time sequence data of the single-tree breast diameter and the corresponding investigation time interval based on the continuous investigation data of a fixed time interval, thereby reversely pushing the age of the single tree in the natural forest stand and providing a convenient method for acquiring the age of the tree. Disclosure of Invention The invention provides a method for determining the age of single wood in a natural forest stand aiming at the defects of the prior art. In order to achieve the above object, the present invention adopts the following technical scheme: a method for determining the age of single wood in natural forest stand includes such steps as choosing the reserved wood with multi-stage actual breast diameter, classifying the tree by annual growth rate of breast diameter, constructing the dummy variable of tree class, using the start point constraint, building the statistical model of multi-stage breast diameter and relative investigation time interval, and back-pushing the age of single wood. Further, the method for determining the age of the single wood in the natural forest stand comprises the following steps: step one, fitting the relation between the breast diameters of the multi-period single wood and the corresponding investigation time intervals, and determining model parameters: based on a theoretical growth equation, a relation model of multi-stage single wood breast diameters and corresponding investigation time intervals is established, wherein the total number of the single wood breast diameter growth models is 2, and the form is as follows: The first model is based on MITSCHERLICH growth equation, adopts starting point constraint, and reflects tree classification on the parameter A, and the model is as follows: And a second model, namely taking a Richards growth equation as a basic model, adopting a starting point constraint, and simultaneously embodying tree classification on the parameters A and C, wherein the model is as follows: Wherein, the A=a1·S1+(a1+d1)·S2+(a1+2d1)·S3 (3) C=c1·S1+(c1+g1)·S2+(c1+2g1)·S3 (4) Wherein D ij is the actual measured breast diameter of the i-th plant of reserved wood in the j-th period, m is the investigation period number, D i1 is the actual measured breast diameter of the i-th plant of reserved wood in the investigation initial period, ad j is the time interval between the j-th period and the investigation initial period, a 1、b、c1 is a model param