Geotechnologies in mapping on land use and cover in agricultural watershed | Geotecnologias no mapeamento do uso da terra e cobertura na bacia hidrográfica agrícola

Resumo

Through the use of geotechnologies many surveys can be performed with data processing as geographical and environmental variables, resulting in new information and products and providing a detailed study about an specific area. This paper’s objective is present a case study example where visual classification appears as an opportunity for mapping heterogeneous areas, especially in agricultural zones occupied by small rural properties classified as family. This mapping type presents advantages to the automatic classifiers, since the heterogeneity of small areas can cause pixels miscellany, damaging the classification results. This study was performed in Salto do Lontra’s watershed, located in Paraná’s southwestern, Brazil. Images from Google Earth, obtained in 2014, were used for a visual classification, with 8 regions of interest (ROIs). For a supervised classification Landsat-8 images were used, path/row 223/78, acquired in 02/12/2013. The classifier applied was Spectral Angle Mapper (SAM) with 5 ROIs. Using such type of technologies, what is expected is that these can provide a broader view over environmental reality, allowing the recognition of different surface features and enabling a correct land management based on adequate information.