Detecting Regime Shifts in the Mean and Variance: Methods and Specific Examples Sergei Rodionov JISAO, University of Washington, Seattle, WA. JISAO, University of Washington, Seattle, WA.
% significance level RSI 1910 Detecting Shifts in the Mean January PDO Searching for the first regime shift RSI – Regime Shift Index l = 10
% significance level RSI Searching for the next regime shift January PDO l = 10
Entry Form
l = p = 0.1 The North Pacific Index (Nov-Mar) RSI
p = 0.05 l = Arctic Oscillation,
Effect of Outliers
PDOa PDOw PDOs PDOa PDOs PDOa PDOs PDOw ALPI NPI NCAR PNA NPI CPC NPI NCAR PDOw AO EPI PDOs EPI AI NPI CPC Regime Shifts in Climatic Indices p = 10 l = 0.1
Std = 1 Std = 2
1989
Conclusions Characteristics of the sequential method: Characteristics of the sequential method: –Automatic detection of regime shifts, –Improved performance at the ends of time series, –Can be tuned up to detect regimes of different scales, –Can handle the incoming data regardless of whether they are presented in the form of anomalies or absolute values, –Works well with the time series containing a trend, –Can be applied to a large set of variables.