This study presents the application of different neural models to a real-life problem of studying the environmental conditions of Castilla y León region (Spain). The goal of this work is to visually and intuitively analyze the level of air pollution in four points of this Spanish region between years 2007 and 2014. The analyzed data were provided by four data acquisition stations from the regional control network of air quality. The main pollutants measured by these stations are analyzed in order to study how the geographical location of these stations and the different seasons of the year are decisive in the behavior of air pollution. Different models for supervised and unsupervised dimensionality reduction have been applied, and subsequent interesting conclusions are obtained.