|
Activity Number:
|
201
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
| Abstract - #307279 |
|
Title:
|
Trend Estimation in a Periodic Survey of North-American Waterfowl
|
|
Author(s):
|
Mark Otto*+
|
|
Companies:
|
U.S. Fish and Wildlife Service
|
|
Address:
|
11510 American Holly Drive, Laurel, MD, 20708-4002,
|
|
Keywords:
|
REGCMPNT models ; ARIMA ; time series ; signal extraction
|
|
Abstract:
|
We estimate trends in North American duck and goose populations from May Breeding Waterfowl and Habitat Survey estimates. The MBWHS is an aerial transect survey systematically placed transects in each of 70 strata encompassing the U.S. Dakotas and Montana, Canada, up to the Northwest Territories and tundra in Alaska. Transects are sampled each year. The trends in the ``true values'' are estimated using the level of the first differenced series or the linear slope in the series itself. The "true values"' are the observed survey estimates apart from the survey sampling error. The survey sampling error is defined from sampling error covariances. Usually periodic surveys only account for sample variance at each time point. The change between time period estimates can be improved by accounting the sampling error covariances among time points.
|
