Open AccessRESEARCH ON TERRITORIAL DIFFERENTIZATION OF TREE-RING CHRONOLOGIES OF ORDINARY PINE BASED ON CLUSTER ANALYSIS
Author(s) -
А. Н. Хох,
Вячеслав Борисович Звягинцев
Publication year - 2018
Publication title -
teorìâ ta praktika sudovoï ekspertizi ì krimìnalìstiki
Language(s) - English
Resource type - Journals
eISSN - 2708-5171
pISSN - 1993-0917
DOI - 10.32353/khrife.2018.41
Subject(s) - dendrochronology , tree (set theory) , cluster (spacecraft) , statistics , homogeneous , ranking (information retrieval) , mathematics , computer science , geography , artificial intelligence , combinatorics , archaeology , programming language
Attempt to evaluate the potential of using dendrochronological information in combination with methods of multidimensional statistical analysis to establish regional belonging of pine lumber is presented. Calculations of the radial increment parameters were performed using the automated workplace DendroExp (accuracy to 0.01 mm). As a result of the research, 85 generalized standardized tree-ring chronologies were constructed. Standardization was carried out separately for each tree followed by averaging of radial increase index according to the object (i. e., on temporary trial areas) by of a simple 5-year moving average method. Subsequently, a gap of 60 years was considered in received chronologies. The calculated statistical parameters such as synchronization factor, coefficient of sensitivity, etc., confirmed the fact that all the analyzed tree-ring chronologies are representative and therefore can be used in dendroecological researches. To answer the question if there are some groups/regions within the Republic of Belarus that similar in dynamics of radial increase have using cluster analysis at the program Statistica 10.0. As a result eight clusters of homogeneous tree-ring chronologies of pine similar in dynamics of radial growth were allocated. Thus, using proposed cluster analysis algorithm a large array of quantitative dendrochronological data can be analyzed which along with other characteristics of tree-ring chronologies (general trends of graphs, expressed periods of oppression, correlation coefficients and synchronism, etc.); it will allow to solve tasks with a high degree of reliability identifying place of growth of felled pine lumber and confirmation of the declared place of its harvesting.