The formation of urban heat islands is one of the effects related to urbanization, as it directly affects thermal comfort. There are several methodologies for its estimation, we can highlight the Gaussian (the best known), whose limitation focuses on the analysis of Gaussian surfaces. When the surface presents cases of poly-nucleated conglomerates, another type of approach (non-Gaussian) is recommended, such as the Quantile method. Therefore, this work seeks to estimate the intensity of surface urban heat islands (SUHI) in the long term (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016) with both methodologies (Gaussians and Quantiles). Based on satellite data and the Weather Research and Forecasting (WRF) meteorological simulation, both with a special resolution of 5 km, for the metropolitan areas of Belo Horizonte, Manaus and Salvador, located in Brazil. Both methods indicate the formation of intense heat islands in the hottest months in the 3 cities studied, with less monthly variation compared to the surface temperature of the Earth's surface.