Time series clustering is an important field of data mining and can be used to identify interesting patterns.This study introduces a new way to obtain clusters of time series by representing them with feature vectors that define the trend, seasonality and noise components of each series, in order to identify areas of the Iberian Peninsula that follow the same pattern of change in their maximum temperature during 1931-2009.Singular spectrum analysis decomposition in a sequential manner is used for dimensionality reduction, which allows the extraction of the trend, seasonality and residual components of each time series corresponding to an area of the Iberian region; then, the feature vectors of the time series are obtained by modelling the extracted components and estimating the parameters.Finally, the series are clustered using a clustering algorithm, and the clusters are defined according to the centroids.The results identified three differentiated zones, allowing to describe how the maximum temperature varied: in the north and central zones, an increase in temperature was noted over time, and in the south, a slight decrease, moreover different seasonal variations were noted according to zones.