A calculation method of stand mean height and DBH considering neighborhood and geographical difference

Abstract

The invention discloses a calculation method of stand mean height and DBH considering neighborhood and geographical difference, which comprises the following steps: data extraction, regional area statistics, establishment of layered data sets, model training, model precision correction, model calculation and checking of the whole data. The method provides a modeling process for tree species with historical survey data which lack growth models or for which the existing growth models are not applicable based on the growth environment, tree species, operation status and various kinds of survey data, and applies spatial domain statistics, spatial interpolation, geographically weighted regression, machine learning and ArcGIS software for building models. The calculation method of average tree height and average DBH of dominant tree species in regional subcompartment, which can self-learn and optimize according to the data, can dynamically establish the calculation model for the target subcompartment to calculate the average tree height and DBH of the subcompartment in the regional scope, and can provide reliable data support for the calculation of forest stock.

Description

technical field [0001] The invention relates to the technical field of stand growth research, and specifically relates to a calculation method for stand average height and DBH in consideration of neighborhood and geographical differences. Background technique [0002] In recent years, the number and scope of stand growth models have reached an unprecedented level. From the early standard harvest table to the current single tree growth model, a continuous model continuum has gradually formed. Due to the different purposes of building models, the mathematical methods for constructing models are also different. Burkhart, Avery, and Tang Shouzheng divided the stand growth models into the following three categories according to the scale of the model and the estimated results: the whole stand model based on the overall stand characteristic index variables (which can be divided into variable density model and average density model). stand growth model), the diameter-order distri...

AI Technical Summary

Problems solved by technology

[0004] Due to the different purposes of building models, the mathematical methods for constructing models are also different. Most of the traditional methods use statistical methods that rely on mathematical equations and a large number of sample data. Debugging ability

With the development of machine learning, more and more researchers have introduced it into the study of stand diameter distribution, but the existing models and methods also have the following defects: firstly, the number and types of existing models are various, but almost All models do not consider the impact of other management levels (such as fertilization, tending, etc.) and the growth law of the stand itself on the growth of the stand except for thinning; secondly, most of the current research is on the common tree species of management (such as fir, masson pine, larch, etc.), the modeling process requires a large number of samples and sample data, and the collection of data requires a lot of labor and material costs, and it is difficult to cover all areas and tree species. Due to the geographical limitations of trees, the established model cannot be easily applied to other regions and tree species; in addition, because the growth of forest trees has strong temporal and spatial characteristics, it is necessary to establish a model suitable for the spatial and temporal characteristics of the target area, while the existing Most of the models in the previous sample plots and wood data (20, 30, and 50 years ago) started from the overall characteristics, and established a model to estimate the overall value, while ignoring the individual differences of the small class, and the model lacks self-correction and design investigation Data matching and adaptive capabilities

Images