Enhancing the precision of broad-scale forestland removals estimates with small area estimation techniques

Abstract National Forest Inventories (NFI) are designed to produce unbiased estimates of forest parameters at a variety scales. These include means and totals current area volume, as well components change such growth harvest removals. Over the last several decades, there has been steadily increasing demand for smaller geographic areas and/or finer temporal resolutions. However, sampling intensities many NFI reliance on design-based estimators often leads inadequate precision these This research focuses improving removal both in terms spatial resolution through use small estimation techniques (SAE). In this application, Landsat-derived tree cover loss product information from mill surveys were used auxiliary data area-level SAE. Results southeastern US suggest improvements can be realized when using make relatively fine Specifically, estimated volume by species group size class was improved SAE methods employed over post-stratified, alone. The findings have broad implications analysts or users interested providing with increased scales than those generally supported post-stratified estimators.