Stratifying to Reduce Bias Caused by High Nonresponse Rates: A Case Study from New Mexico's Forest Inventory.

摘要

The USDA Forest Service's Forest Inventory and Analysis (FIA) Program applies specific sampling and analysis procedures to estimate a variety of forest attributes. FIA's Interior West region uses post-stratification, where strata consist of forest/nonforest polygons based on MODIS imagery, and assumes that nonresponse plots are distributed at random across each stratum. The sample of New Mexico's forests during 2005-2013 did not meet this assumption due to a large number of private landowners who denied permission to access plots on their properties. We developed a modified stratification scheme to minimize bias in our estimates of forest attributes. The purpose of this paper is twofold: first, to document the customized stratification methods used for the New Mexico forest inventory of 2005-2013, and second, to provide sufficient instructions for application of this method to other states or inventories where high nonresponse rates occur.