Prediction of daily modes of South Asian monsoon variability and its association with Indian and Pacific Ocean SST in the NCEP CFS V2

Intra‐seasonal variability of the South Asian monsoon and its relationship with the Indo–Pacific sea‐surface temperature in the NCEP CFSv2

“…The larger lag window is usually employed with the MSSA whereas EEOF analysis includes smaller lag window which is a key difference between MSSA and EEOF. The MSSA method extracts the spatial structure of oscillatory and non‐oscillatory modes and has been applied to study the ISO in the climate physics (Plaut and Vautard, ; Krishnamurthy and Shukla, , ; Achuthavarier and Krishnamurthy, , ; Krishnamurthy et al ., ; Shahi et al ., ).…”

“…It is also observed that from P2 lead onwards, model produces stationary modes and the northwest to southeast tilt also significantly reduce. The spatial structure of the 42‐day oscillatory mode obtained in the present analysis is similar to that of the 45‐day mode seen in the observation (Shahi et al , , figure 3). The significant improvement in the space–time structure of the oscillatory mode in the model produced precipitation is observed in current forecasts especially for P1 and P2 leads as compared to the previous forecasts from the CFSv2 model (Shahi et al , ).…”

“…There are many previous studies in which multi‐channel singular spectrum analysis (MSSA) method proved to be a robust technique to distinguish intra‐seasonal oscillatory modes and persisting modes from daily anomalies of precipitation data over South Asian monsoon region (Krishnamurthy and Shukla, ; Krishnamurthy and Shukla, ; Achuthavarier and Krishnamurthy, , ; Krishnamurthy et al ., ; Shahi et al ., ). The improvement of prediction in terms of period of oscillation, space–time structure, and the propagation of oscillatory modes is observed in National Centres for Environmental Prediction (NCEP) climate forecast system version 2 (CFSv2) model implemented by India for extended range prediction (Roxy et al ., ; Sahai et al ., ; Sharmila et al ., ; Saha et al ., ; Liu et al ., ; Shahi et al ., ) as compared to previous version of the same model (Seo et al ., ; Yang et al ., ; Wang et al ., ; Achuthavarier and Krishnamurthy, ; Lee Drbohlav and Krishnamurthy, ; Pattanaik and Kumar, ; Achuthavarier and Krishnamurthy, ; Chaudhari et al ., ; Suhas et al ., ; Saha et al ., ). However, there is still hope for the improvement in the CFSv2 model so that the forecast may be at par with observation.…”

“…However, there is still hope for the improvement in the CFSv2 model so that the forecast may be at par with observation. It was also observed that seasonally persistent modes are related to El Niño‐Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) (Krishnamurthy and Kirtman, ; Achuthavarier and Krishnamurthy, , ; Shahi et al ., ) but these relationships were not robust in the forecast models. The signature of dependency of seasonally persistent modes with external forcing such as ENSO is seen in the CFSv1 and CFSv2 (Achuthavarier and Krishnamurthy, ; Shahi et al ., ) but it is not clearly distinguishable as in the case of observations.…”

“…The signature of dependency of seasonally persistent modes with external forcing such as ENSO is seen in the CFSv1 and CFSv2 (Achuthavarier and Krishnamurthy, ; Shahi et al ., ) but it is not clearly distinguishable as in the case of observations. However, the variations of ENSO and IOD occur in conjunction with each other in the CFSv1 (Achuthavarier and Krishnamurthy, ) as well as in CFSv2 (Shahi et al ., ). The teleconnection of ENSO and Indian summer monsoon rainfall (ISMR) is more realistic in the CFSv2 as compared to CFSv1 (Saha et al ., ).…”