Health conditions of Eucalyptus plantations in Brazil

ArcelorMittal runs Eucalyptus plantations in Brazil for the production of charcoal that is used in steel production facilities. ArcelorMittal Florestas, a company belonging to AM, is managing some four sites in the regions of Bahia and Minas Gerais. The forest managers are interested in the potential of EO data to support their forest monitoring, especially with regards to the identification of the health status and the growing conditions within the plantations. The service should deliver EO based indicators highlighting forest conditions and health status during the crucial period, which is the dry season between June and September. EO data at three points in time should be acquired to enable the monitoring. In addition to that, the images need to be georectified and a land cover map produced based on one image.
Executive Summary (extracted from the Service Operation Report prepared by the service provider GIM)

The overall objective of the project is an assessment of the suitability of EO-derived information in support of the forest monitoring programme, focussing on indicators of forest conditions and health status, as well as spatial variability within the different stands. The basic idea of the project is thus to use adapted vegetation indices and colour composites to depict the spatial heterogeneity in the different stands in order to target the field work and to provide information for remedial actions (e.g. chemicals application or replanting). As the most critical period for forest plantations is the winter dry season, new satellite images have to be acquired from June to September. A time series of very high resolution satellite images (SPOT sensor, 10m multispectral) over the vegetation period were acquired, i.e. 19/07/2008, 10/08/2008 and 30/07/2008 covering the Centre West Plantation (only AM Florestas property) north of Bom Despacho in Minas Gerais.

In addition to land cover maps of the region, a number of Vegetation Indices (VIs) were computed. The best results were achieved by mapping inhomogeneities over the forest stands. The resulting maps show locations, where the VIs are different from the expectations so these stands should be checked on the field for potential damages.

Assessment of impact and benefit by the user


  • The service provided us with useful information on the potential of EO data for the management of the plantation.
  • Bearing in mind the size of the plantation, the use of EO data saves us a lot of field work by directing the forest workers to the exact location where the trees have problems in their development.
  • Rapid identification of problems avoids large damage of the trees and the enlargement of eventual diseases that would result in large monetary losses.
  • The precision and quality of the data is very good.


  • The delay in receiving the deliveries (especially the change detection products) should be decreased, because the faster eventual problems are identified, the less the damage will be.
  • The time delay between two images shall be shorter to increase the potential to detect anomalies.