A method for mapping corn using the US Geological Survey 1992 National Land Cover Dataset

Computers and Electronics in Agriculture
By: , and 



Long-term exposure to elevated nitrate levels in community drinking water supplies has been associated with an elevated risk of several cancers including non-Hodgkin's lymphoma, colon cancer, and bladder cancer. To estimate human exposure to nitrate, specific crop type information is needed as fertilizer application rates vary widely by crop type. Corn requires the highest application of nitrogen fertilizer of crops grown in the Midwest US. We developed a method to refine the US Geological Survey National Land Cover Dataset (NLCD) (including map and original Landsat images) to distinguish corn from other crops. Overall average agreement between the resulting corn and other row crops class and ground reference data was 0.79 kappa coefficient with individual Landsat images ranging from 0.46 to 0.93 kappa. The highest accuracies occurred in Regions where corn was the single dominant crop (greater than 80.0%) and the crop vegetation conditions at the time of image acquisition were optimum for separation of corn from all other crops. Factors that resulted in lower accuracies included the accuracy of the NLCD map, accuracy of corn areal estimates, crop mixture, crop condition at the time of Landsat overpass, and Landsat scene anomalies.

Publication type Article
Publication Subtype Journal Article
Title A method for mapping corn using the US Geological Survey 1992 National Land Cover Dataset
Series title Computers and Electronics in Agriculture
DOI 10.1016/j.compag.2005.11.003
Volume 51
Issue 1-2
Year Published 2006
Language English
Publisher Elsevier
Contributing office(s) Earth Resources Observation and Science (EROS) Center
Description 12 p.
First page 54
Last page 65
Google Analytic Metrics Metrics page
Additional publication details