Warming SST of Bay of Bengal and decrease in formation of cyclonic disturbances over the Indian region during southwest monsoon season

Abstract

Low-pressure system (LPS), a major rain-bearing synoptic circulation, forming over the Indian region, including Bay of Bengal and Arabian Sea plays a vital role in performance of southwest monsoon over the country. The term LPS includes lows, depressions and cyclonic storms. According to the intensities, LPS are categorized into two, one only low-pressure areas (LPA) and the other more intense systems like depressions/storms (DDS). Statistical analysis reveals some significant results. Decadal analysis shows that there is a significant increase(decrease) in the frequency and duration of LPA(DDS) during the monsoon season for the recent decades. SST of Bay of Bengal also increased significantly during recent period. It is also observed that frequency and duration of LPA(DDS) show significant positive(negative) trend and sea surface temperature (SST) of the Bay of Bengal shows significant positive trend for the period after 1960. The total frequency of LPS has neither increased nor decreased significantly but the duration of LPS has significantly increased. This means, while the average total formation of the systems remains the same, the duration has increased. It seems that there are some atmospheric and oceanic conditions which are responsible for not allowing the intensification of lows into depressions. The frequency and duration of LPA(DDS) during the monsoon season are positively(negatively) correlated with SSTs of the Bay of Bengal during winter, pre-monsoon and monsoon season indicating warmer SST of the Bay of Bengal may not be favourable for intensifying lows into depressions.

Appendix

How, the test applied, is briefly explained in the following. Let x series have m terms and y series n terms. Mix-up the terms of the two series and arrange the combined series in ascending order of values, and separately add up the ranks of the terms of the x series and the ranks of the terms of the y series in the combined series in the ascending order.

Let Rx and Ry be sums of the ranks of term of the x and y series respectively in the combined series. Let Rx be smaller than Ry. The test statistic is U where

$${\text{U}} = {\text{Rx}} - \left[ {{{{\text{m}}\left( {{\text{m}} + 1} \right)} \mathord{\left/{\vphantom {{{\text{m}}\left( {{\text{m}} + 1} \right)} 2}} \right.\kern-\nulldelimiterspace} 2}} \right]$$

U is Gaussian or near-Gaussian for m,n>12.

For m,n≤12, values of U significant at the desired levels may be obtained from the relevant tables (Fix and Hodges 1955).

For m,n>12,E(U)=mn/2,${\text{Var}}\;{\text{U}} = {{{\text{mn}}\left( {{\text{m}} + {\text{n}} + 1} \right)} \mathord{\left/ {\vphantom {{{\text{mn}}\left( {{\text{m}} + {\text{n}} + 1} \right)} {12}}} \right. \kern-\nulldelimiterspace} {12}}$

Deviation of U from the expected mean in terms of the standard deviation (SD) of ${\text{U}} = \frac{{\left[ {{\text{Rx}} - \frac{{{\text{m}}\left( {{\text{m}} + 1} \right)}}{2} - \frac{{{\text{mn}}}}{2}} \right]}}{{\sqrt {\frac{{{\text{mn}}\left( {{\text{m}} + {\text{n}} + 1} \right)}}{{12}}} }}$

Standardized deviation of U may be referred to tables of standard Gaussian distribution for judging the significance of U. If U is significant, the difference in the levels of the two groups of values is significant.