Biennial Oscillation in Temperature and Monsoon Activity

Mohanakumar K, V. Sathiyamoorthy and S. Sijikumar

• Introduction
• Data and Methodology
• Results and discussion
• Constant Phase of Temperature TBO with Height
• Link between QBO and TBO in Temperature
• QBO/TBO in Zonal Wind
• Conclusion
• References

FIGURES

• Figure 1
• Figure 2
• Figure 3
• Figure 4
• Figure 5
• Figure 6
• Figure 7

Abstract

            Signals of biennial oscillations with periods ranging from 20-32 months are noted in the tropospheric temperature over the near-equatorial Indian station, Thumba (8°23¹ N, 76°52¹ E). The phase of the Tropospheric Biennial Oscillation (TBO) in temperature does not vary with height from surface to the level of tropopause and is found to be associated with the intensity of the monsoon rainfall. Temperature over Thumba shows Quasi Biennial Oscillation (QBO) in lower stratosphere. Phases of the QBO and TBO in temperature meet at tropopause level. Where they meet, phases of the QBO and TBO are unsynchronized during the decade 1971-1981 and synchronized during next decade, 1982-1992. The QBO in zonal wind has neither inter- decadal variability nor disturbances.

Introduction

Quasi Biennial Oscillation (QBO) in zonal wind is a dominant natural oscillation in the equatorial lower stratosphere. The driving force for the QBO is the vertical transfer of momentum from the troposphere to stratosphere by Kelvin and Rossby-Gravity waves. One of the current problem is the vertical coupling of the QBO to the mesosphere in the up, and to the troposphere below, its influence on the tropospheric weather conditions etc. Several studies have revealed the biennial type signals in meteorological quantities in the troposphere (Angell and Korshover,1964; Terray, 1995; Trenberth, 1975 etc). Biennial oscillation is found in the rainfall over India (Mooley and Parthasarathy, 1984), Indonesia (Yasunari and Suppiah, 1988). A strong biennial component is seen in the ENSO phenomena (Rasmusson et al, 1990; Roplewski et al, 1992) and is proved to be an integral part of the Asia-Pacific climate. This Tropospheric Biennial Oscillation (TBO) has an irregular period of 2-3 years and shows eastward movement.

The cause for the TBO is not clearly known. Several theories and hypothesizes have been proposed to address the cause, development, evolution of TBO. Majority of the theories suggest that the TBO is caused by the interaction of atmosphere-ocean-monsoon systems and extra tropics (Nicholls, 1978; Meehl, 1994, 1997; Chang and Li, 2000, etc). In an entirely different mechanism, TBO is considered as the result of the forcing from the QBO in the lower stratosphere through an unclear stratosphere ­ troposphere interactions. Gray et al (1992) suggested a mechanism by which QBO can alter the deep convective activity over west Pacific. Khandekar (1996, 1998) proposed a conceptual model by which QBO can produce a biennial oscillation in Indian monsoon activity. Using General Circulation Model simulations, Giorgetta et al (1999) investigated the mechanisms by which the QBO could influence the tropical tropospheric circulation in areas of deep convections. Several studies showed that the phase of the QBO influence the weather events in the troposphere. Indian summer monsoon rainfall activity is seems to be directly related with the phase of the QBO (Bhalme et al, 1987, Mukherjee et al, 1985). Gray (1984) identified an association between QBO phase and seasonal incidence of Atlantic Tropical cyclones.

Using zonal wind data from Singapore, Korur and Ponape, Yasunari (1989) suggested that there exists a coherent phase structure between the lower stratospheric zonal wind QBO and biennial oscillations scale zonal wind anomalies in the lower and upper troposphere. He also showed that the QBO in the lower stratosphere is also coupled with that in the sea surface temperature anomalies in the equatorial Pacific. Yasunari¹s (1989) results are based on the inferences from the results obtained from only one parameter (zonal wind) over a few stations in the maritime continent and equatorial Pacific regions. It is interesting to know whether this kind of phase coherence is appearing between TBO and QBO over Indian monsoon region and if it is so, whether it is present in other parameter also?  The station selected for this purpose Thumba, is under the influence of both summer and winter monsoons of the Indian peninsula and tropical lower stratospheric QBO.

Data and Methodology

Radiosonde measured monthly mean temperature and zonal wind data at 1km interval for the altitude range 1-27 km during September 1971 to December 1992 over Thumba (Trivandrum) constitute the basic data for the present study. Beyond this period, the radiosonde data are not available in this station. Temperature anomalies of each month from the 256 month mean were computed for each level. Data of Indian Summer Monsoon Rainfall (ISMR) has been taken from Parthasarathy et al (1994). ISMR is the average June to September rainfall of 306 stations well distributed over India.

To analyse the multi-time scale oscillations present in zonal wind and temperature anomaly, Morelet wavelet transform is used. It is a useful tool to analyse the time series that contain non-stationary power at many different frequencies (Daubechies, 1990). Wavelet transform transforms a one dimensional time series into two-dimensional frequency-time domain. So it is possible to know the frequency content of the signal at every time-step. Wavelet transform uses generalized wave functions called wavelets that can be stretched and translated both in time and frequency.

Wavelet decomposes a signal s(t) in terms of some elementary function ?_b,a (t) derived from a analyzing wavelet or mother wavelet ?(t) (Weng and Lau, 1994). The wavelet transform of a real signal s(t) with respect to the analyzing wavelet ?(t) may be defined as

            ?^*        is the complex conjucate of ?

            b          is the position (translation)

            a(>0)    is dilation.

The analyzing wavelet ?(t) for Morelet wavelet transform is ?(t)=e^ik?t  e^-?txt?/2, which is a plane wave modulated by a Gaussian. Complex Morelet wavelet transform was applied to zonal wind and temperature at each level. This transform provides information of signal on both amplitude and phase. Length of the data is kept as 256 (2⁸) in order to avoid edge effects caused by data padding and voices per octave are set as 4. Zero octave was fixed as 8 months. Detailed mathematical treatment of the wavelet transform is available elsewhere (Lau and Weng, 1995; Torrence et al, 1998; Weng and Lau, 2000).

Results and discussion

In fig. 1, the time series¹ of zonal wind and temperature anomaly at two levels, 11 and 21 km representing middle troposphere and lower stratosphere are presented. Corresponding real part of the Morelet wavelet transform is presented in fig 2 & 3. In zonal wind, annual mode is the only oscillation in the troposphere and both annual and biennial oscillations are strong in the lower stratosphere. In temperature anomaly, 11-year solar cycle, biennial, annual and semi annual oscillations are seen in both lower stratosphere and troposphere (octaves -2, -1, 0, 1, 2, 3, 4 corresponds to 2, 4, 8, 16, 32, 64, 128 month periodicities respectively). This clearly shows the presence of biennial oscillations both in lower stratosphere and troposphere in temperature and only in lower stratosphere in zonal wind.

Fig. 1

Fig. 2 Real part of the Morelet wavelet coefficients of monthly mean temperature anomaly at 11 km

Fig. 3 Real part of the Morelet wavelet coefficients of monthly mean temperature anomaly at 21 km

Constant Phase of Temperature TBO with Height

Wavelet analysis shows prominent TBO with periods ranging from 20-32 months in temperature anomaly at almost all levels of troposphere over Thumba. Time sries of real part of the temperature anomaly mean wavelet coefficients corresponding to the biennial mode (20-32 months) for the troposphere and lower stratosphere with 2km uniform interval is presented in fig. 4. This is similar to filtering the biennial mode and removing other modes using standard filtering techniques. The time-height plot of the real part of the mean wavelet coefficients of temperature anomaly corresponding to the biennial mode is presented in fig. 5. Real part of the wavelet coefficients provides information about both the phase and intensity of a signal at a given time and scale relative other times and scales (Weng and Lau, 1994). Amplitude of TBO varies with time and height. TBO is found to be strong during 1971-81 in the troposphere, whereas it is relatively weak during 1982-92. It is clear from figs. 4-5, that the phase of TBO in temperature anomaly from surface to tropopause is nearly constant throughout the study period.

Fig. 5 Time-height plot of  wavelet filtered temperature anomaly corresponding to biennial mode.

Link between QBO and TBO in Temperature

Wavelet analysis of temperature anomaly shows QBO in lower stratosphere. Phases of QBO propagate downward with time in the 27-24 km altitudes during 1971-81. Downward phase propagation is disturbed and becomes of constant phase with height at 23-18 km altitudes during the same period. Downward phase propagation of QBO is weak or absent during 1982-92. Phases of QBO and TBO meet at tropopause level or just above 15 km altitude. Where they meet, phases of QBO and TBO are unsynchronized during 1971-81. On the other hand, during 1982-92, phases of QBO and TBO are synchronized there and none of the unusual phenomena of 1971-81 period is manifested.

To get a quantitative idea about the link between QBO and TBO in temperature, the real part of the mean wavelet coefficients of the biennial mode at all levels have been grouped into first (September 1971-December 1981) and second (January 1982-December 1992) decades. Since the QBO signal is prominent at 22 km altitude, QBO at this level is taken as reference QBO and correlation coefficients have been worked out between the real part of the 22 km biennial mode mean wavelet coefficients and those of other levels (1-27 km) for the entire study period and also for the first and second decades separately. The correlation coefficient values are presented in fig. 6.

Correlation coefficients of first and second decades show marked differences. In the first decade, correlation coefficient is positive in the 23-15 km levels and negative above 23 km and below 15 km levels. Opposite signs of correlation in 23-15 km and below 15 km levels clearly show that QBO and TBO phases are unsynchronized during the first decade. Opposite signs between 27-24 km and 23-15 km altitudes show the sudden disturbances in the downward propagating phase of QBO at 23-15 km altitudes. In upper and lower troposphere, magnitudes of correlation are high and it is lowest at 9 km. In the second decade, correlation coefficients are positive at all the levels except tropopause. This shows the existence of phase synchronization between TBO and QBO during this period. For the entire period the correlation show a similar pattern as that of the first decade.

QBO/TBO in Zonal Wind

Wavelet analysis is applied to zonal wind of Thumba for the same 256 month period as that of temperature to study the characteristics of QBO/TBO in it and compare them with those of temperature. In lower stratosphere, clear and regular QBO signal is seen in zonal wind (fig. 7). Phase of the zonal wind QBO propagates downward with uniform phase speed of about one km/month without any disturbance at any level. No significant variation is noticed in the maximum easterly and westerly amplitudes. Zonal wind QBO does not show marked difference between the two decades unlike in temperature.

Fig. 7  Time-height plot of  wavelet filtered zonal wind corresponding to biennial mode.

Zonal wind hardly shows any TBO signal in the troposphere throughout the study period. As explained by Yasunari (1989), weaker TBO signal over Thumba in zonal wind is due to the location of Thumba near a node or small amplitude area of the wave no 1 or 2 structure of TBO mode around the globe.

Link between Monsoon Activity and Phase of QBO/TBO in Temperature and zonal wind

Thumba, located in the southern most part of the Indian peninsula is under strong influence of the Asian summer monsoon circulation. Our analysis shows a significant TBO signal in tropospheric temperature and prominent QBO signal in stratospheric temperature as well as zonal wind over Thumba. Recent studies reveal the existence of biennial time scale variability in Indian summer monsoon activity. According to Meehl (1987, 1997), Indian summer monsoon activity is strongly modulated by biennial time scale variability caused by atmosphere-ocean coupled processes occurring in the Indian and Pacific oceans. Convective activity is strong (weak) over the Indian ocean area (represented by Thumba data) during strong (weak) Indian monsoons. Kanamitsu and Krishnamurti (1978) pointed out a major shift in the circulation patterns from their normal position over Indian to southeastward during a drought year.

Table.1 gives data regarding ISMR for 1972-92. Of these years 1972, 1979, 1982 and 1987 are the weakest monsoons and 1975, 1983 and 1988 are

Table 1. Indian Summer Monsoon Rainfall 1972-1992 (Parthasarathy et al. 1994)

the strongest monsoons of the period. During the 3 strong Indian monsoons the temperature TBO is in positive phase. Generally stratospheric zonal wind QBO is in westerly phase during strong Indian monsoon and easterly phase during weak Indian monsoons as pointed out by Bhalme et al, (1987). Of the 4 weak Indian summer monsoons except in 1979 TBO is in negative phase. The results suggest that the observed biennial variability in the tropospheric temperature over Thumba, may be due to the monsoon-ocean-atmosphere interactions taking place over Indian ocean region in biennial time scales as suggested by Meehl (1997).

Conclusion

We have examined the vertical phase structure of TBO and QBO in Temperature over the near equatorial Indian station Thuma. The decadal change of phase coherence between TBO and QBO in temperature and correlation between these oscillations suggests that they are different phenomena. Two nearly biennial phenomena with slightly different periods are expected to drift in and out of phase on decadal timescales. The strong relation between the phase of TBO and Indian summer monsoon activity suggest that the atmosphere-ocean-monsoon interaction may be the cause for the TBO.

References

Angell J. K, and J. Korschover, 1964: J. Atmos. Sci., 21, 479-492.

Bhalme, H. N., et al, 1987: J. Climatol., 7, 345-353.

Chang, C. P and T. Li, 2000: J. Atmos. Sci, 57, 2209-2224.

Daubechies, I, 1990: IEEE Trans. Inform. Theory, 36, 961-1004.

Giorgetta, M. A, et al, 1999: Climate. Dyn, 15, 435-450.

Gray, W. M., et al, 1992: Geophy. Res. Lett., 19, 107-110.

Kanamitsu, M and T. N. Krishnamurti, 1978: Mon. Wea. Rev, 106, 331-347.

Khandekar, K. L., 1998: J. Climate, 11, 3057-3059.

Khandekar, M. L., 1996: Kluwer Academic, 76-95.

Lau K. M and H. Y. Weng, 1995: Bull. Amer. Met. Soc., 76, 2391-2402.

Meehl, G. A., 1994: Science, 266, 263-267.

Meehl, G. A., 1997: J. Climate, 10, 1921-1943.

Mooley, D. A and B. Parthasarathy, 1984: Mon. Wea. Rev., 111, 967-978.

Mukherjee, B. K., et al., 1985: Mon. Wea. Rev., 113, 1421-1429.

Nicholls, N, 1978: Mon. Wea. Rev., 106, 1505-1508.

Parthasarathy, B et al, 1994: Theor. Appl. Climatol., 49, 217-224.

Rasmusson, E. M et al, 1990: J. Mar. Sci, 1, 71-96.

Ropelewski, C. F et al, 1992: J. Climate, 5, 594-614.

Terray, P. 1995: J. Climate, 8, 2595-2618.

Torrence, C and G. P. Compo, 1998: Bykk Amer. Met. Soc., 79, 61-78.

Trenberth, K. E., 1975: Quart. J. Roy. Met. Soc, 101, 55-74.

Weng, H. Y and K. M. Lau, 1994: J. Atmos. Sci., 51, 2523-2541.

Yasunari, T and R. Suppiah, 1988: Deepak Publishing, Hampton, pp-527.

Yasunari, T, 1989: J. Met. Soc. Japan, 67, 483-493.