Close Window

MODIS-based estimates of row-crop agricultural expansion in Rond˘nia and Mato Grosso

Gillian Laura Galford, Brown University; MBL, (Presenting)
John Frasier Mustard, Brown University,
Jerry Michael Melillo, MBL,
Carlos Eduardo Cerri, ESALQ/USP,
Carlos Clemente Cerri, CENA/USP,
Shannon M Pelkey, Brown University,

Rapid changes in land use in the southwestern Amazon, particularly conversion of native vegetation to mechanized row-crop agriculture, have many ecological consequences, including changes in biogeochemical cycles. Quantifying the patterns and magnitude of changes in row-crop agriculture is a first step in estimating regional-scale environmental impacts. We used five years of MODIS data to identify annual changes in row-crop agriculture, focusing on expansion and intensification of croplands in Mato Grosso and Rond˘nia. Our approach involved a wavelet-smoothing methodology for processing and analyzing MODIS time series data. We checked our results against crop histories from farms in both states. From annual statistics of phenology, we distinguished areas of cropland from other land covers. Results show increases in area of row-crop agriculture of 1% in Rond˘nia and 6% in Mato Grosso between 2001-2005. Within row-crop agriculture, we calculate the timing and number of crops per year (single or double cropping). Between 2001 and 2005, farmers in Mato Grosso have intensified from single to double cropping on more than 11,000 square kilometers, while in Rond˘nia, farmers have transitioned to double cropping on over 20 square kilometers. As next steps, we will modify the Terrestrial Ecosystems Model (TEM) to incorporate these new remote sensing results and then estimate the impacts of land use change on greenhouse- gas emissions, particularly nitrous oxide.

Science Theme:  LC (Land Use and Land Cover Change)

Presentation Type:  Poster

Abstract ID: 27

Close Window