Introduction

A great wealth of resources has been expended to inventory our Nation’s forests, and an equally substantial amount of effort has gone into acquiring remotely sensed data. As such, these two data types comprise the ends of a continuum of detail. Plot inventory data are extremely sparse geographically but have a high level of information content regarding the resources at the inventory plot locations. Conversely, remotely sensed data cover the entire globe but with comparatively limited information at any single location. The common approach to leverage these two forms of data has been to create thematic maps of vegetation-related classes as well as response surfaces for other target variables of interest. With regard to vegetation mapping, these thematic maps typically describe dominant vegetation and include physiognomic, floristic, or structural characteristics or all. These thematic maps often include some additional land use classes or map land use as a separate theme. The variables available for analysis are limited to the classes included in the map, and once the analysis is complete there is no ability to develop new attributes without a new mapping effort. Recently, a more flexible approach, single neighbor imputation, has been utilized to provide sample tree lists and plot calculated variables for all the unsampled pixels in raster maps. Although not replacing traditional mapping methods, imputed maps can greatly enhance analytical flexibility and provide information in a familiar context that is often supported by extensive simulation modeling capability. Imputed maps are not intended to suggest that each pixel is in fact occupied by the imputed plot data, but rather given what current information I have, what do I expect. However, developing and mapping 30-m products over broad spatial extents is a lengthy process. This project was conducted in order to evaluate the differences in using a new national spatial predictor database at a coarser resolution (250 m) instead of the 30-m data used in the previous study.