Title: Misclassification bias in areal estimates
Author: Czaplewski, Raymond L.
Source: Photogrammetric Engineering and Remote Sensing. 58(2): 189-192
Description: In addition to thematic maps, remote sensing provides estimates of area in different thematic categories. Areal estimates are frequently used for resource inventories, management planning, and assessment analyses. Misclassification causes bias in these statistical areal estimates. For example, if a small percentage of a common cover type is misclassified as a rare cover type, then the area occupied by the rare type can be severely overestimated. Many categories are rare in detailed classification systems. I present an informal method to anticipate the approximate magnitude of this bias in statistical areal estimates, before a remote sensing study is conducted. If the anticipated magnitude is unacceptable, then statistical calibration methods should be used to produce unbiased areal estimates. I then discuss existing statistical methods that calibrate for misclassification bias with a sample of reference plots.
Keywords: remote sensing; forest inventories; misclassification; bias
View and Print this Publication (500 KB)
- We recommend that you also print this page and attach it to the printout of the article, to retain the full citation information.
- This article was written and prepared by U.S. Government employees on official time, and is therefore in the public domain.
Get the latest version of the Adobe Acrobat reader or Acrobat Reader for Windows with Search and Accessibility
Czaplewski, Raymond L. 1992. Misclassification bias in areal estimates. Photogrammetric Engineering and Remote Sensing. 58(2): 189-192.