Geostatistic Applied to Spatial Modeling of Hypsometric Relationships in Forest Stands
Allan Libanio Pelissari, Afonso Figueiredo Filho, Sidney Fernando Caldeira, Sebastião do Amaral Machado

Abstract
Height trees can be estimated more realistically if the structure of spatial dependence is considered in hypsometric relationships. Thus, the aim of this study was to apply geostatistical techniques to modeling the spatial patterns and estimate total height in teak stands. Average values of total height (H) and diameter at 1.3 m above ground (DBH) were obtained from 273 sample units at the second and sixth years, after selective thinning. Four models were fit using traditional hypsometric relationships. Also, geostatistical analyses were used to model the spatial patterns of height, as well as cross-semivariograms to estimate the height when correlated with DBH. Traditional modeling was more accurate than simple geostatistical analysis; however, including the spatial variability of DBH, results were statistically superiors to map the height. Cross-semivariogram and cokriging analyses identified the spatial correlation of the height with DBH and described the spatial variability of young and post-thinning stands.

Full Text: PDF