The objective of this paper was to test the application of geo-statistical models to estimate the volume of wood timber residues in a forest area managed in Paragominas, state of Pará. The Warren and Olsen's and Ordinary Kriging methods were used to estimate the volume of forest residues in sampled and not sampled points. The results confirm that the applied methods were appropriate, given that the estimates for the volume of wood timber residues, in the area of study, showed a small difference of 5.2%, when compared to the actual value obtained.
Amazon forest; forest management; geo-statistics