Sampling Type Influence in the Estimation of Eucalyptus Wood Volume by Kriging

ABSTRACT The objective was to verify if the estimate of the individual volume of wood per tree conducted by kriging is influenced by the type of sampling. A settlement with 1875 eucalyptus trees was marked on a Cartesian plane (X, Y) and strictly cubed using the Smalian method. Six samples were taken, each containing 106 trees. Two samples per railing, two random samples and two by crowded grid. The semivariograms were constructed for each of the sampling and maps of the individual volume of trees were constructed by kriging. Eleven types of errors were compared and ranked. Random sampling showed the best results and clustered sampling grid showed the worst results.