Title: Practical Considerations When Using Perturbed Forest Inventory Plot Locations To Develop Spatial Models: A Case Study

Author: Coulston, John W.; Reams, Gregory A.; McRoberts, Ronald E.; Smith, William D.

Year: 2006

Publication: In: Proceedings of the sixth annual forest inventory and analysis symposium; 2004 September 21-24; Denver, CO. Gen. Tech. Rep. WO-70. Washington, DC: U.S. Department of Agriculture Forest Service. 126p.

Abstract: U.S. Department of Agriculture Forest Service Forest Inventory and Analysis plot information is used in many capacities including timber inventories, forest health assessments, and environmental risk analyses. With few exceptions, actual plot locations cannot be revealed to the general public. The public does, however, have access to perturbed plot coordinates. The influence of perturbed plot coordinates on the development of spatial models is unknown. We examined the influence by comparing the accuracies of two spatial models for predicting forest biomass, ordinary kriging and residual kriging. We developed each model using the actual coordinates and 10 independent perturbations of the actual coordinates. We tested for differences in accuracy using analysis of variance. No statistically significant difference in accuracy was found. The results represent only a small portion of the possible outcomes, however. We suggest a simulation study to examine the spatial range of influence that plot coordinate perturbation has on model accuracy.