A principal component analysis based model to predict post-monsoon tropical cyclone activity in the Bay of Bengal using oceanic Nino index and dipole mode index

作者：Biswas H. R.^*; Kundu P. K.

来源：International Journal of Climatology, 2018, 38(5): 2415-2422.

DOI：10.1002/joc.5344

• tropical cyclone
• Bay of Bengal
• ENSO
• ACE
• PCR model

El Nino Southern Oscillation (ENSO) and Indian Ocean dipole (IOD) play an important role in determining the weather pattern over the Indian subcontinent region. The role of ENSO and IOD on the occurrence of tropical cyclone activity during post-monsoon season (October-December) over the Bay of Bengal (BoB) has been investigated through an objective analysis of observed data for the period of 1990-2015. Accumulated cyclone energy (ACE) index is an important measure of tropical cyclone activity over a basin for a defined period of time. Sea-surface temperature (SST) anomalies in the Nino 3.4 region (oceanic Nino index) is negatively correlated at 95% confidence level of significance with ACE over the BoB during post-monsoon season. Positive phase of IOD has negative impact on the tropical cyclone formation over the BoB. The ACE has large inter-annual variability with coefficient of variation 124% for tropical cyclone activity over BoB during post-monsoon season. Frequency distribution of annual ACE values for post-monsoon season over the BoB indicated that cold phases of ENSO along with negative IOD index values are the most favourable for development of tropical cyclone over the BoB. Principal component regression (PCR) model developed by cross-verification method based on training period data for 1990-2013 and ONI and DMI values of different lag periods as predictors is found to be functional for both deterministic and probabilistic prediction of ACE values of post-monsoon season. Above- and below-normal TC activities were observed in 2014 and 2015, respectively, which have been well predicted by the PCR model.

• 出版日期2018-4