Quantifying predictability of Indian summer monsoon intraseasonal oscillations using nonlinear time series analysis
The chaotic time series analysis is used to estimate the predictability of the Indian summer monsoon intraseasonal oscillations (ISOs). An index of the ISO is constructed using the daily gridded high-resolution (0.25 x 0.25 ) Indian summer monsoon rainfall dataset during 1951-2007. A low-dimensional chaotic attractor for the monsoon ISO with a correlation dimension of 3.8 is identified for a saturation embedding dimension of 5. The correlation dimension gives the minimum number of variables, whereas, the saturation embedding dimension gives the maximum number of variables needed to describe the state of the system. Thus, a minimum of 4 variables and a maximum of 5 variables should be able to simulate the ISO. The largest Lyapunov exponent of the ISO index time series is 0.05. The limit of predictability of the monsoon ISO is nearly 3 weeks.