“…This approach is useful for vegetation studies, especially in agricultural areas, since vegetation cover is quite dynamic in time, and the ability to capture these variations is essential to discriminate different types of crops, through its phenological characteristics. For this purpose, most studies have explored time series of vegetation indices like NDVI or EVI from MODIS data (Jakubauskas et al, 2002;Sakamoto et al, 2005;Wardlow et al, 2007;Esquerdo et al, 2011;Arvor et al, 2011;Körting, 2012;Risso et al, 2012;Coutinho et al, 2013;Borges & Sano, 2014;Tomás et al, 2015;Neves et al, 2016). However, there is a demand to produce more detailed maps, which can be obtained from higher spatial resolution satellites such as Landsat-like (Zheng et al, 2015;Peña et al, 2015;Pan et al, 2015).…”