This paper is looking to show to use of system data collected from wide-area monitoring systems (WAMS). They allow monitoring of the dynamics of power systems. Among the WAMS applications, there is the modal identification algorithm, which identifies critical oscillatory modes from PMU measurements. This application permits using data processors for estimating of frequency, damping, and amplitude of dominant mode oscillations observable in a specific electric signal (e.g., active power, frequency) recorded for the analyzed period. However, since modal identification of real-time measurements is based on an online optimization, the results usually have considerable fluctuations. Thus, it is essential to consider the complementary implementation of trend analysis for acquiring convenient early-warning indicators of oscillatory problems. This consideration allows avoiding erroneous information of the systems oscillatory behavior of the system real-time that modal identification of crude results could deliver. In this paper, the application of a l1 filter for determining the trend analysis of high-dimensional data set resulted from a commercial modal identification is explored. The algorithm is applied to an oscillatory event registered by the WAMS of the Ecuadorian National Interconnected System with promising results.