LAUSR

Observational and modeling studies have identified an El Niño Southern Oscillation (ENSO) like Ocean-Atmospheric coupled phenomenon in the tropical Atlantic during the boreal summer season, popularly known as Atlantic Zonal Mode (AZM). The atmospheric teleconnection between the AZM and Indian Summer Monsoon Rainfall (ISMR) is significant especially in non-ENSO years. Hence, realistic simulation of AZM and its teleconnection with ISMR in coupled Ocean-Atmospheric models is important in improving the predictive skill of ISMR. Here, we analyzed the outputs of nine-month hindcast simulations of a coupled ocean-atmospheric model (NCEP CFSv2) in estimating the prediction skill of AZM and its teleconnection with ISMR. It is found that the AZM prediction skill in CFSv2 hindcasts initialized in the month of February (FebIC) is poor and it simulates an unrealistic teleconnection with ISMR. The CFSv2 FebIC hindcasts are unable to forecast the correct phase of the low level wind anomalies over western Atlantic basin, which leads to an error in the forecast of AZM. It is shown here that the error in the prediction of AZM and its erroneous teleconnection with ISMR leads to a loss of prediction skill of ISMR in CFSv2 FebIC hindcasts. It is further noted that the shorter lead time forecast of CFSv2 initialized in the month of May (MayIC hindcasts) shows improved prediction skill of AZM and its associated teleconnection with ISMR compared to the longer lead time forecast (FebIC hindcasts). However, the prediction skill of ISMR is low in MayIC hindcasts compared to FebIC hindcasts. We find that the low prediction skill of ISMR in May IC hindcasts is due to the misrepresentation of the teleconnection between ENSO and ISMR. A notable improvement in the prediction skill of ISMR in CFSv2 can be achieved by combining the ENSO induced ISMR from FebIC hindcasts and AZM induced ISMR from MayIC hindcasts. These results aid in improving the ISMR seasonal forecast in CFSv2. Besides, this study highlights the need to improve the Atlantic variability and its teleconnection with ISMR in the pursuit of improving prediction skill of ISMR in coupled climate models.