The Amazon has been observed mainly through the deforestation phenomenon, using
traditional resources from remote sensing as quantification of deforested area and its
subsequent annual increasing, which sounds to be an effective methodology. Confirming this
argument, I noted in a survey with 16.591 environmental fees applied by IBAMA/PA in the
period 2000 to 2008, that more than 85,0% were related with biodiversity (flora), and
specifically in the year of 2008 at the jurisdiction of IBAMA Santarém (west of Pará), almost
60% of fees had linkage with deforestation, identified by remote sensing. Analysis using only
satellite images are not enough to identify elements of surface, with a poor contribution to the
understanding and intervention of reality. It were interviewed and vectorized479 rural
properties in the region of Paragominas and Santarém, state of Pará, which have historical
land usedistinguished, qualifying them by technological trajectories in the Costa`s sense,
correcting distortions of economic development, aggregating information at remote sensing
data.It was applied geotechnologic resources of landscape metrics and were built a cell
database integrated with statistics and algorithms of probabilistic optimization, associating
classification non supervised isodata (validated with kappa index = 0,87, considered
“excellent”) with production types collected in the field, generating an integrated
classification (kappa index = 0,78, considered "very good"). In Paragominas were qualified 3
different types of technological trajectories, trajectory peasant.T8 (dominated by temporary
culture), peasant.T3 and employer.T4 (specialized in beef cattle). In the region of Santarém,
appeared trajectory peasant.T2 (high presence of permanent and temporary culture and agro
forestry systems) and employer.T7 (mutation of T4, enhancing the participation of temporary
cultures). The methodology was considered successful, spatializing rural properties by
technological trajectories and generating land use classes more adequated, as temporary
culture and pasture, but in the isodata classification was only “agropecuary”, resulting a more
realistic viewof production activities developed in the investigated area, materializing the
generation of spatial information by integrating data from different sources and increasing the
capability of pixel analysis.
|
