GRL-TBO

Tropospheric biennial oscillation over the southeast Pacific intertropical convergence zone during ENSO breaks

Chih-wen Hung and Michio Yanai
Department of Atmospheric Sciences, University of California, Los Angeles
Los Angeles, California
Submitted to Geophysical Research Letters
September 2001

Abstract. The tropospheric biennial oscillation (TBO) observed in the southeast Pacific intertropical convergence zone (SITCZ) during two ENSO breaks (1979-82 and 1994-97) is studied. The SITCZ occurs only in March and April in the region (130°W-90°W, 8°S-2°S). Biennial fluctuations are observed in satellite-measured precipitation rate, surface wind convergence and integrated cloud liquid water of these months. A possible link of the TBO in the SITCZ to the biennial variability of rainfall in northeastern Brazil is suggested.

1. Introduction
Biennial signals in rainfall, sea-level pressure, sea-surface temperature (SST) and other elements in the tropical Indian and western Pacific regions [the tropospheric biennial oscillation (TBO)] have been studied by many authors [e.g., Trenberth, 1975; Meehl, 1987, 1993; Yasunari, 1991; Shen and Lau, 1995; Webster et al. 1998; Chang and Li, 2000]. In the eastern Pacific, however, the TBO signal has been difficult to isolate from the more dominant El Niño events.
Over the eastern Pacific Ocean, the Intertropical Convergence Zone (ITCZ) is a prominent feature in the atmospheric general circulation. The ITCZ over the eastern Pacific prefers the Northern Hemisphere, resulting from the ocean-atmosphere interactions and coastal geometries [e. g., Philander et al., 1996]. It reaches the northernmost location (20°N) in June, July and August, and then returns to the lower latitude (5°N) in February, March and April. In addition to this major ITCZ, a band of convergence zone, the southeast Pacific intertropical convergence zone (denoted SITCZ), appears in March and April in the Southern Hemisphere [Halpern and Hung, 2001; hereafter HH01]. They made a detailed study of the SITCZ using satellite observations (1993-98) and detected biennial oscillations in precipitation and surface wind convergence. In the present study, the TBO in the SITCZ is studied further with additional observations, and its possible link to the biennial variability of rainfall in northeastern Brazil is examined.

2. Data
The monthly data used for the period of 1979-98 are: (1) Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP) [Xie and Arkin, 1997] and (2) Niño 3 (150°W - 90°W, 5°S - 5°N) sea surface temperature (SST) anomalies provided by CPC ( http://www.cpc.noaa.gov/data/indices). For the period of 1993-98, (1) European Remote-sensing Satellites (ERS) 10-meter wind [Offiler, 1994], (2) Special Sensor Microwave/Imager (SSM/I) integrated cloud liquid water [Wentz, 1997], and (3) Reynolds SST provided by the National Centers for Environmental Predictions (NCEP) [Reynolds and Smith, 1994] are used.

3. SITCZ and ENSO
The 20-year average of CMAP data shows the double ITCZ over the eastern Pacific only in March and April (Figure 1a). Recently, Zhang [2001] studied the global double ITCZs and found the most remarkable double ITCZ in this region in these two months. On the other hand, the September-October average of CMAP data showed only one ITCZ as in all other months (Figure 1b). Following HH01, we choose a (130°W - 90°W, 8°S - 2°S) box shown in Figure 1 as the "SITCZ region". In Figure 2, the time series of the monthly precipitation rate for 1979-98 in the SITCZ region is shown together with Niño 3 SST anomalies. The monthly precipitation rate exhibits distinct annual cycles with the maximum in every March and April, and almost no rainfall between July and December.
In Figure 2, abnormally high precipitation rates, more than two times of the normal SITCZ value (20-year March-April average, 3.94 mm day-1), are observed in 1983 and 1998. Both years correspond to strong El Niño events, for which the anomalies of Niño 3 SST index were higher than 3°C. The abnormal rainfall rates in these two years were caused by migrations of the ITCZ across the equator from the Northern Hemisphere to the Southern Hemisphere (Figure 3). This time-longitude figure shows the ITCZ migration of this type occurred in strong El Niño years (1983 and 1998) as well as in weak El Niño years (1987, 1992, and 1993).
In Figure 2, we note that the rainfall rates in the SITCZ during La Nina years (1984, 1985, 1986 and 1989) are also larger than the normal SITCZ value. It is possible that the enhanced cold tongue in these years resulted a larger meridional gradient of SST and this created stronger low-level convergence [as suggested by Lindzen and Nigam, 1987] in SITCZ area. In summary, the heavy March-April rainfall in SITCZ region in 1983, 1987, 1992, 1993 and 1998 was associated with the El Niño, and that in 1984, 1985, 1986 and 1989 was related to the La Nina.

4. TBO in the SITCZ during ENSO breaks
Excluding the El Niño and La Nina events, there are two periods (1979-82 and 1994-97) which are long enough to include 4 normal years. In these ENSO breaks, biennial fluctuations of precipitation rate are observed (Figure 2). The March-April rainfall rates in even years (1980, 1982, 1994 and 1996) were larger than those in odd years (1979, 1981, 1995 and 1997). Further evidence of the TBO in the 1994-97 break is obtained from the horizontal divergence of the ERS 10-meter wind (Figure 4a) and integrated cloud liquid water (Figure 4b). In both figures, the abnormally large values in 1998 are due to the El Niño.
In Figure 4a, the March-April values of surface convergence in the even years, 1994 and 1996, are 4.22x10-6 and 4.50x10-6 s-1, respectively, which have the mean, 4.36x10-6 s-1. In the odd years, 1995 and 1997, the March-April surface convergence values are 2.75x10-6 and 1.74x10-6 s-1, respectively, with the mean value 2.24x10-6 s-1. The biennial range of surface wind convergence, the March-April difference between the even years (strong TBO years) and the odd years (weak TBO years) is 2.12x10-6 s-1, which is larger than the root-mean-square (RMS) accuracy, 1.5x10-6 s-1 [HH01]. A similar method can be used to examine the TBO of integrated cloud liquid water. The mean of March-April integrated cloud liquid water in 1994 and 1996 is 101.48 g m-2, and that in 1995 and 1997 is 85.20 g m-2. The biennial range is 16.28 g m-2, which is larger than the 10% uncertainty of the measurement [Ferraro et al., 1996]. These independent satellite observations are in phase with each other, which increase the credibility of the TBO observed in the SITCZ.
Although the annual cycle of SST (Figure 4c) plays an important role in producing the SITCZ in every March and April, the influence of SST on the TBO in the SITCZ is not clear. The biennial range of March-April SST (0.03°C) in 1994-97 obtained from Figure 4c is much less than the RMS accuracy, 0.5°C [Reynolds and Smith, 1994]; therefore, the local air-sea interaction theory [e.g. Meehl, 1993] may not explain the TBO in the SITCZ.

5. The connection of TBO in the SITCZ to that in northeastern Brazil
The March-April average difference of precipitation rate between the strong TBO years (1980, 1982, 1994, 1996) and the weak TBO years (1979, 1981, 1995, 1997) shows the same sign in the SITCZ and northeastern Brazil, and the opposite sign over the cold tongue area north of the SITCZ (Figure 5). The time series of monthly precipitation rate over northeastern Brazil (60°W - 40°W, 2.5°S - 5°N; shown in Figure 5) for 1979-98 shows suppression of rainfall during the El Niño events as well as biennial fluctuations during the two ENSO breaks (Figure 6). We suspect that the TBO in the SITCZ has a connection with that in the northeastern Brazil.
A possible scenario for this connection is that an east-west circulation transmits the TBO signal from northeastern Brazil to the eastern Pacific cold tongue area. When northeastern Brazil is in a strong TBO year, the enhanced downward motion over the cold tongue will increase the southward surface flow along the northern edge of the SITCZ, and enhance the convergence in the SITCZ. During a weak TBO year, the opposite mechanism will reduce the convergence in the SITCZ. The ERS surface wind observation tends to support this hypothesis (Figure 7).
At present, the mechanism of the TBO in northeastern Brazil is unknown. However, the Atlantic ITCZ strongly influences the rainfall in northeastern Brazil [Hastenrath and Heller, 1977]. The TBO in the equatorial Atlantic Ocean was studied by using a coupled atmosphere-ocean general circulation model [Tseng and Mechoso, 2001]. They concluded that the Atlantic TBO can be generated by coupled atmosphere-ocean interactions internal to the Atlantic Ocean itself independent of the ENSO in the Pacific. On the other hand, Chiang et al. [2000] showed the observational evidence that Pacific and Atlantic ITCZ variabilities are linked through an east-west Walker circulation. Saravanan and Chang [2000] used an atmospheric general circulation model to demonstrate the interaction between ENSO and tropical Atlantic variability through the same east-west circulation. Although these two studies did not particularly emphasize the TBO, a similar mechanism through the east-west Walker circulation may explain the TBO connection between the SITCZ and northeastern Brazil.

6. Summary and discussion
During two ENSO breaks (1979-82, and 1994-97), biennial fluctuations are observed in the precipitation rate in the SITCZ (Figure 2). For the 1994-97 break, independent satellite-based measurements show biennial signals also in the surface wind convergence and integrated cloud liquid water (Figure 4).
Comparisons of the TBO signals in precipitation rates in the SITCZ region and northeastern Brazil (Figures 2, 5, and 6) suggest a link between the two regions through an east-west vertical circulation. Examination of these figures leads us to infer the following: (1) During El Niño/La Nina events, the eastern Pacific strongly influences the east-west circulation, and the northeastern Brazil rainfall is suppressed/enhanced as a passive response. (2) During ENSO breaks, however, the TBO in northeastern Brazil occurs without ENSO influence, and the signal is transmitted to the eastern Pacific cold tongue and resulting the TBO in the SITCZ.
The TBO in the SITCZ suggests a mechanism in which the east-west vertical circulation between the Pacific and Atlantic Oceans is not always dominated by the ENSO. During the ENSO breaks, the signal such as TBO in the Atlantic and northeastern Brazil can be transmitted to the eastern Pacific.

Acknowledgments. This research was supported by NOAA Grant NA96GP0331 and by NSF Grant ATM-9902838. Special thanks to Dr. David Halpern of JPL for his encouragement and providing the ERS data for this study. The help of P. Woiceshyn of JPL for processing the ERS data and the assistance given by Dr. Dennis Shea and NCAR NCL group are appreciated.

References
Chang, C.-P. and T. Li, A theory for the tropical tropospheric biennial oscillation, J. Atmos. Sci., 57, 2209-2224, 2000.
Chiang, J. C. H., Y. Kushnir, S.E. Zebiak, Interdecadal changes in eastern Pacific ITCZ variability and its influence on the Atlantic ITCZ, Geophys. Res. Lett.., 27, 3687-3690, 2000.
Ferraro, R. R., F. Weng, N. C. Grody, and A. Basist, An eight-year (1987-1994) time series of rainfall, clouds, water vapor, snow cover, and sea ice derived from SSM/I measurements, Bull. Amer. Meteorol. Soc., 77, 891-905, 1996.
Halpern, D., and C.-W. Hung, Satellite observations of the southeast Pacific Intertropical Convergence Zone During 1993-1998, J. Geophys. Res., accepted, 2001.
Hastenrath, S., and L. Heller, Dynamics of climate hazards in northeast Brazil, Quart. J. R. Meteorol. Soc., 103, 77-92, 1977.
Lindzen, R. S., and S. Nigam, On the role of sea surface temperature gradients in forcing low-level winds and convergence in the tropic, J. Atmos. Sci., 44, 2418-2436, 1987.
Meehl, G. A., The annual cycle and interannual variability in the tropical Indian and Pacific Ocean regions, Mon. Wea. Rev., 115, 27-50, 1987.
-------, A coupled air-sea biennial mechanism in the tropical Indian and Pacific regions - role of the ocean, J. Climate, 6, 31-41, 1993.
Offiler, D., The calibration of ERS-1 satellite scatterometer winds, J. Atmos. Oceanic Technology, 11, 1002-1017, 1994.
Philander, S. G. H., D. Gu, D. Halpern, G. Lambert, N. -C. Lau, T. Li, and R. C. Pacanowsky. Why the ITCZ is mostly north of the equator, J. Climate, 9, 2958-2972, 1996.
Reynolds, R. W., and T. M. Smith, Improved global sea surface temperature analyses using optimum interpolation, J. Climate, 7, 929-948, 1994.
Saravanan, R. and P. Chang, Interaction between tropical Atlantic variability and El Nino-Southern Oscillation, J. Climate, 13, 2177-2194, 2000.
Shen, S., and K. M. Lau, Biennial oscillation associated with the east Asian summer monsoon and tropical sea surface temperature, J. Meteor. Soc. Japan, 73, 105-124, 1995.
Trenberth, K. E., A quasi-biennial standing wave in the Southern Hemisphere and interrelations with sea surface temperature, Q. J. R. Meteorol. Soc., 101, 55-74, 1975.
Tseng, L. S., and C. R. Mechoso, A quasi-biennial oscillation in the equatorial Atlantic Ocean, Geophys. Res. Lett.., 28, 187-190, 2001.
Wentz, F. J., A well-calibrated ocean algorithm for special sensor microwave/imager, J. Geophys. Res., 102, 8703-8718, 1997.
Webster, P. J., V. O. Magana, T. N. Palmer, J. Shukla, R. A. Tomas, M. Yanai, and T. Yasunari, Monsoons: Processes, predictability, and the prospects for prediction, J. Geophys. Res., 103, 14451-14510, 1998.
Xie, P. P. and P. A. Arkin, Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs, Bull. Am. Meteorol. Soc., 78, 2539-2558, 1997.
Yasumari, T., The monsoon year - a new concept of the climatic year in the tropics, Bull. Am. Meteorol. Soc., 72, 1331-1338, 1991.
Zhang, C. D., Double ITCZs, J. Geophys. Res., 106, 11785-11792, 2001.

Figures

Figures in JPG format: Figure 1-2 Figure 3-4 Figure 5-7

Figure 1. The 1979-98 average precipitation rate for (a) March-April and (b) September-October. Contour interval is (mm day -1). A rectangular box shows the SITCZ region (130°W - 90°W, 8°S - 2°S).
Figure 2. Time series of monthly precipitation rate (mm day -1 ) averaged over the SITCZ region from 1979 to 1998. The gray horizontal line represents the 20-year mean precipitation rate (3.94 mm day -1) in March and April in this region. Also shown by a thick line is the monthly Niño 3 SST anomaly (°C ) .
Figure 3. Latitude-time section for the monthly precipitation rate averaged between 130°W - 90°W. Contour interval is 2 mm day -1.
Figure 4. Time series of monthly mean values averaged over the SITCZ region from 1993 to 1998: (a) surface wind divergence (10 -6 s -1 ), (b) integrated cloud liquid water (g m -2 ), and (c) SST (°C).
Figure 5. Difference of the mean March-April precipitation rate [the average for strong TBO years (1980, 1982, 1994, 1996) minus the average for weak TBO years (1979, 1981, 1995, 1997)]. Contour interval is 1 mm day -1. Two rectangular boxes indicate the SITCZ (130°W - 90°W, 8°S - 2°S) and northeastern Brazil (60°W - 40°W, 2.5°S - 5°N) regions.
Figure 6. Similar to Figure 2, but for the northeastern Brazil region (see Figure 5). The gray horizontal line represents 20-year mean precipitation rate (8.91 mm day -1) in March and April in this region. Also shown by a thick line is the monthly Niño 3 SST anomaly (°C ) .
Figure 7. Difference of the mean March-April surface wind [the average for strong TBO years (1994, 1996) minus the average for weak TBO years (1995, 1997)]. Also shown is the corresponding difference of the mean precipitation rate (mm day -1).

Updated on Nov 19 2001
*Corresponding author address: Dr. Michio Yanai, Department of Atmospheric Sciences, University of California, Los Angeles, 405 Hilgard Avenue, Los Angeles, CA 90095-1565 Email: yanai@atmos.ucla.edu