This paper describes a comparative study in which a procedure using ground gathered data and classified LANDSAT data for estimating crop area was compared to procedures using ground gathered data and sampled LANDSAT data. Data from parts of Iowa and Missouri were used. Unitemporal data were used in Iowa and multitemporal data were used in Missouri. Results indicate that some sampling schemes can be used without any significant difference in the crop area estimates, but with a large reduction in cost for corn, soybeans, and winter wheat.

Date of this Version