This paper describes an approach in which multitemporal Landsat-5 Thematic Mapper (TM) data from the pre-planting season, when the cloud cover is minimal, was used in order to evaluate the area to be planted with annual crops in the rainy season. Both supervised and non-supervised classification of multitemporal Normalized Difference Vegetation Index (NDVI) images were evaluated. The interpretation of the areas to be planted with summer crops was based on field work, NDVI temporal profiles, unitemporal classifications, and the crop calendar of the region. A correction based on the error matrix was applied to the classification and the results were compared to the estimation provided by the official agencies using kappa statistics analysis. The proposed method was shown to be a feasible alternative for improving the crop estimation programme, especially for the tropical countries where the cloud cover has restricted the use of optical remote sensing data.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
No Article Media
Document Type: Research Article
National Institute for Space Research (INPE), Remote Sensing Division, P.O. Box 515, 12201-970, São José dos Campos, SP, Brazil; gabriela,epiphani, Email: [email protected]
April 1, 2003
More about this publication?