Revista Brasileira de Cartografia (Dec 2005)

Soybean crop area estimation in Rio Grande do Sul through Landsat images

  • Rodrigo Rizzi,
  • Bernardo F. T. Rudorff

Journal volume & issue
Vol. 57, no. 3
pp. 226 – 234

Abstract

Read online

The objective of this work was to map and estimate soybean crop area in municipalities of Rio Grande do Sul State, Southern Brazil, through digital classification and visual interpretation based on multitemporal images acquired during the combined operation of Landsat-5 and -7 satellites, in crop year 2000/01. Six scenes from paths 220 to 224, rows 79 and 80 were acquired at two key dates to identify soybean crop areas. The soybean crop map was carried out through digital classification followed by a meticulous visual interpretation of the multitemporal images in order to correctly map all soybean areas based on its typical temporal-spectral behavior. Crop area was estimated based on the soybean map and compared with official estimation provided by the Brazilian Geography and Statistics Institute (IBGE). The largest relative differences were observed in municipalities where soybean plantation is less expressive, although greatest absolute differences were observed in municipalities with large amount of soybean (above 10.000 ha). For the overall State estimate a difference of 11.3% between IBGE (2.773.498 ha) and Landsat images (2.492.880 ha) was observed. The particular simultaneous operation condition of Landsat-5 e -7 satellites during the crop year of 2000/01 showed that cloud free images acquired at key periods can be used to correctly map soybean areas and to assist the official agricultural statistics improvement.

Keywords