Chapter 6 6.2 Statistical methods time series analysis methods (2)Autocorrelation: temporal autocorrelation e.g. today's wind speed is similar to yesterdays wind speed this is the persistence or inertia of the climate system Diurnal/annual cycles should be removed before estimating autocorrelation X-average (X)/ fitting and removing an analytical function such as a series of sinusoids or polynomials The most natural way to visualize autocorrelation in a time series is by plotting the autocorrelation as a function of lag time. that is the autocorrelation function Assessment of Climate/Change ImpactsChapter 6 Assessment of Climate/Change Impacts 6.2 Statistical Methods • time series analysis methods (2) Autocorrelation: temporal autocorrelation • e.g. today’s wind speed is similar to yesterday’s wind speed this is the persistence or inertia of the climate system • Diurnal/annual cycles should be removed before estimating autocorrelation • X-average(X)/ fitting and removing an analytical function, such as a series of sinusoids or polynomials • The most natural way to visualize autocorrelation in a time series is by plotting the autocorrelation as a function of lag time, that is the autocorrelation function