Publication

Abstract : This study proposes an efficient framework employing Ensemble Empirical Mode Decomposition (EEMD) algorithm coupled with Time-Dependent Intrinsic Cross Correlation (TDICC) method to detect the teleconnection between large scale climatic oscillations and monthly rainfall of India. Indian Ocean Dipole (IOD), El Niño Southern Oscillation (ENSO) and the enhanced convective phases of Madden Julian Oscillation (MJO) have taken into consideration as climatic oscillations for the teleconnection study. EEMD method decomposed each signal to a set of zero mean oscillatory modes namely Intrinsic Mode Functions (IMFs)with definite periodicity. The time-dependent running correlation analysis of the IMFs of ENSO and IOD showed strong negative correlation with the modes of rainfall, at inter annual scales. All the IMFs of MJO indices 8–10 showed very strong positive correlation while MJO indices 2 and 3 showed very strong negative correlations with the corresponding IMFs of rainfall. TDICC analysis found the most influential antecedent information of climatic oscillation in the prediction of IMFs of rainfall at different time scales. On considering the different lags in TDICC analysis, the high frequency modes are associated with transition in correlation from positive to negative and vice versa while low frequency modes display a stable pattern in the teleconnections of rainfall with climatic oscillations. The TDIC-based identification of relevant modes and TDICC-based identification of significant lags will substantially alleviate the computational complexities and help in improved predictions of monthly rainfall over India.