Monsoon teleconnections over Indian and Pacific Oceans in Japan Meteorological Agency model simulation

摘要

The study evaluates various sea surface temperature (SST) indices and their teleconnections with Indian summer monsoon rainfall (ISMR) simulated by Japan Meteorological Agency (JMA) coupled general circulation model (CGCM). The indices considered for evaluations are Nino indices (Nino 4, Nino 3.4, Nino 3 and Nino 1+2) of Pacific Ocean (PO) and Indian Ocean dipole (IOD) indices [eastern IO (IOD E), western IO (IODW) and dipole mode index (DMI)] of Indian Ocean (IO). The study has been carried out for the season June to September (JJAS) using JMA hindcast SST simulated on various initial conditions May, April, March and February (here after leads 0, 1, 2 and 3, respectively) during the period 1979-2010. The extended reconstructed SST (ERSST) from National Center for Environmental Prediction is used as observed SST. The important outcome of this study is that the inverse relationship between Nino 3.4 and ISMR is found to be re-established in recent years after a period of weakening. The JMA model is able to simulate the same with highest skill on lead 1. The PO SST characteristics including climatology, spatial anomaly pattern and interannual variability of Nino indices are well captured in the model lead 0 simulation (correlation coefficient is above 0.80). The model simulation on lead 0 is able to predict the El Nino Southern oscillation (ENSO) events during hindcast period with good skill. It is also noted from this study that the model skill to simulate PO SST characteristics is decreasing with increase of lead time. This study is also showing that the model has some deficiency to simulate IO SST characteristics. The model performance in the simulation of IOD events is not appreciable. Thus, this study is recommending JMA model as an efficient ENSO predicting tool, however the model requires improvements to predict IOD events.