Spatial and temporal dynamics of land use in the municipality of Seropedica, Rio de Janeiro, Brazil/Dinamica espacial e temporal do uso da terra no municipio de Seropedica, Rio de Janeiro, Brasil

• Published in: Revista geográfica acadêmica 2014-01, Vol.8 (1), p.38
• Main Authors: da Silva, Roberto Diego Bezerra, Delgado, Rafael Coll, Lyra, Gustavo Bastos, Rodrigues, Rafael de Avila, Cardozo, Ana Carolina, Brasileiro, Felipe Gomes
• Format: Article
• Language: Spanish
• ISSN: 1678-7226; 1678-7226

The paper analyses the spatial-temporal dynamics of land use and coverage between 1990 and 2010 in Seropedica, in the Metropolitan Region of Rio de Janeiro (MRRJ). It uses the images captured by the TM (ThematicMapper) Landsat 5 and the supervised method of classification based on the algorithms of Minimum Distance, Maximum Likelihood and Mahalanobis distance. The 1990 results show that the Minimum Distance classifier overestimated the classes of water areas (5.02%), forest (8.94%), exposed soil (16.05%) and pasture (59.33%), respectively. When compared to the other methods, the 1990 results also show that the Minimum Distance classifier underestimated the urban area class (10.66%). In 2010, the Minimum Distance classifier overestimated the classes of pasture (59%), forest (20.68%) and water (2,25%), whereas when compared to the Mahalanobis distance and Maximum Likelihood classifiers the spectral classes of soil exposure (0.92%), urban area (11.54%) and eucalyptus (5.81%) were underestimated. In both classifiers the mean-bias (MB) showed that the values of the area were overestimated in relation to the project SOS Mata Atlantica. The vegetation indices showed to be compatible with current literature. However, the discrepancies in the estimates of area attributed to the transition between dry and rainy seasons, which happened before the satellite had passed over the region, were exceptions. It is important to highlight, however, that all classificatory methods used during this study are liable to error, thus making the work of the photo interpreter indispensable. In other words, these classifiers were developed with the aim to improve the visual interpretation work in Seropedica.