An increased penetration of electric power from intermittent energy sources such as wind raises concerns regarding safe operation of the power system. For system operators and utilities, it is important to understand, predict and control power outputs from individual and multiple Wind Farms (WF). Here the so-called Koopman Mode Decomposition (KMD) is applied to wind turbine outputs, which transforms the time-series data into a finite number of modes evolving with single frequencies. The application enables an intuitive and natural approach to quantification of smoothing effects of wind power outputs.