U.S. flag

An official website of the United States government

NALDC

Main content area

State level crop area estimation using satellite data in a regression estimator


Author:
Graham, M.L.
Source:
USDA
Year:
1993
Subject:
estimation, regression analysis, data collection, mathematical models, and Landsat
Abstract:
The USDA's National Agricultural Statistics Service (NASS) estimates state level crop acreage in the Mississippi Delta region using area frame survey data and Landsat Thematic Mapper (TM) satellite data. Five general steps produce these acreage estimates. First, a sample of TM pixel data is clustered by land cover. Second, sampled TM pixels are assigned to a land cover class using maximum likelihood classification. Third, classified sample pixels are regressed with reported crop acreages. Fourth, TM scenes are classified. Finally, acreage is estimated with a regression estimator using classified pixel counts as ancillary information to the' ground survey' data. The potential· benefit is mainly a reduction in variance with some adjustment of the state acreage estimates.
Language:
English
Permanent URL:
http://handle.nal.usda.gov/10113/21788