|
Activity Number:
|
444
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
JASA, Applications and Case Studies
|
| Abstract - #305074 |
|
Title:
|
Model-Assisted Estimation of Forest Resources with Generalized Additive Models
|
|
Author(s):
|
Jean D. Opsomer*+ and F. Jay Breidt and Gretchen Moisen and Goeran Kauermann
|
|
Companies:
|
Iowa State University and Colorado State University and U.S. Forest Service and Universitaet Bielefeld
|
|
Address:
|
Snedecor Hall, Ames, IA, 50011,
|
|
Keywords:
|
multi-phase survey estimation ; nonparametric regression ; calibration ; systematic sampling ; variance estimation
|
|
Abstract:
|
Multi-phase surveys often are conducted in forest inventories with the goal of estimating forest characteristics over large regions. We describe how design-based estimation of such quantities, based on information gathered during ground visits, can be made more precise by incorporating auxiliary information from remote sensing. The relationship between ground measurements and the remote sensing variables is modeled using generalized additive models. Model-assisted estimators utilizing these nonparametric fits are proposed and applied to forestry survey data from northern Utah. The design context of this study is two-phase systematic sampling from a spatial continuum, and we describe difficulties with the standard variance estimation approach. An alternative assessment of estimator performance based on a synthetic population is discussed.
|
