What is the meaning of UNIT ROOT in time series modelling? I understand that, for a time series data to be stationary, the mean and standard deviation must be constant throughout the period considered. Also, Augmented Dickey Fuller test checks for the presence of unit root in the time series, which if it does, the series is non stationary. I do not understand the intuition behind the presence of unit root and how it relates to a time series' property of being stationary.