Intraseasonal and Interannual Variations of Tropical Convection: A Possible Link between the 40–50 Day Oscillation and ENSO?

Ka-Ming Lau Laboratory for Atmospheres. NASA Goddard Space Flight Center, Greenbelt, Maryland

Search for other papers by Ka-Ming Lau in
Current site
Google Scholar
PubMed
Close
and
Paul H. Chan Applied Research Corporation, 8201 Corporate Drive, Landover, Maryland

Search for other papers by Paul H. Chan in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Intraseasonal and interannual variations of tropical convection are studied using 12 years of NOAA outgoing longwave radiation (OLR) data. The spatial patterns of OLR variance in the 1–5 day, 20–70 day and longer-than-180 day bands am examined separately for the northern summer (May–October) and winter (November&ndash April). In the 1–5 day band, the most prominent features are the Pacific and Atlantic ITCZs during the northern summer and the large variability over the western Pacific and the tropical continents during the northern winter. For the time scales beyond 10 days, it was found that the maximum variance shifts north and south of the equator from summer to winter but is generally located between 20°S and 20°N in the longitudinal sector between the Indian and Pacific oceans. Over the Indian Ocean and the western Pacific, variations with 30–60 day period are most pronounced. In the central and eastern Pacific variations with periods over one year are more prevalent.

The natural variability of tropical convection is estimated. Results show that the climate signal (interannual variability) in convection is 1argest and most easily detectable over the equatorial central Pacific. Results of empirical orthogonal function analyses show that the most dominant mode of variation in tropical convection exhibits multiple time scales that can be identified respectively with the 40–50 day oscillation, the annual cycle and an aperiodic variation with several year' time scale. These variations are associated with fluctuations of the equatorial Walker circulation. It is suggested that an interaction of these time scales may be instrumental in leading to the onset of ENSO. The possibility of the 40–50 day oscillation as a trigger to ENSO is further supported by the seemingly systematic frequency and amplitude modulation of these oscillations before and after the 1982–83 ENSO. Discussion of the results in light of recent development in the dynamics of the coupled ocean-atmosphere system is also presented.

Abstract

Intraseasonal and interannual variations of tropical convection are studied using 12 years of NOAA outgoing longwave radiation (OLR) data. The spatial patterns of OLR variance in the 1–5 day, 20–70 day and longer-than-180 day bands am examined separately for the northern summer (May–October) and winter (November&ndash April). In the 1–5 day band, the most prominent features are the Pacific and Atlantic ITCZs during the northern summer and the large variability over the western Pacific and the tropical continents during the northern winter. For the time scales beyond 10 days, it was found that the maximum variance shifts north and south of the equator from summer to winter but is generally located between 20°S and 20°N in the longitudinal sector between the Indian and Pacific oceans. Over the Indian Ocean and the western Pacific, variations with 30–60 day period are most pronounced. In the central and eastern Pacific variations with periods over one year are more prevalent.

The natural variability of tropical convection is estimated. Results show that the climate signal (interannual variability) in convection is 1argest and most easily detectable over the equatorial central Pacific. Results of empirical orthogonal function analyses show that the most dominant mode of variation in tropical convection exhibits multiple time scales that can be identified respectively with the 40–50 day oscillation, the annual cycle and an aperiodic variation with several year' time scale. These variations are associated with fluctuations of the equatorial Walker circulation. It is suggested that an interaction of these time scales may be instrumental in leading to the onset of ENSO. The possibility of the 40–50 day oscillation as a trigger to ENSO is further supported by the seemingly systematic frequency and amplitude modulation of these oscillations before and after the 1982–83 ENSO. Discussion of the results in light of recent development in the dynamics of the coupled ocean-atmosphere system is also presented.

Save