P2.143 Multidecadal oscillations in tropical ocean and land temperatures

Thursday, 13 May 2010
Arizona Ballroom 7 (JW MArriott Starr Pass Resort)
Carlos D. Hoyos, Georgia Institute of Technology, Atlanta, Georgia; and P. Webster, H. R. Chang, and J. A. Curry

During the last century, there has been a general upward trend in the tropical sea-surface temperatures (SSTs). At the same time, there have been periods of anomalous warming and cooling against this trend. Such anomalous warming periods occurred in the 1935-1945 decade and also in the 1995-2005. The 1940's “bump” is extraordinary because it occurred globally and exists in both the land (station data) and the ocean. Some efforts have attempted to dismiss the bump in terms of uncorrected biases caused by changes in how ocean temperature was measured (bucket versus engine intake). But as the land perturbations match in magnitude (about 0.35oC) almost exactly that which occurred in the ocean, it appears unlikely that the bump wad the result of an uncorrected bias. The common feature between the 1940's bump and the accelerated increase of temperature in the 1995-2005 period was that both the Pacific Decadal Oscillation and the Atlantic Multidecadal Oscillation were positive and in phase.

We investigate the oscillations of tropical ocean and land temperatures during the last 100 years and attempt to extend these analyses (using IPCC model integrations) through the next century. It is found that the area organized and deep convection stays ostensibly constant irrespective of the phase of interdecadal variability or trend. Physical reasons for this constancy are presented. Also, it is found that there are distinct oscillations in the net heating over land and over the oceans and these are generally out of phase. Similar oscillations are found in rainfall (GPCP) over land and oceans. The 1940 bump is consistent with these findings and shows a 4-year lead of maximum temperature over land relative to the ocean. We use these observations to attempt to decipher between natural variation of climate and trends and their relative importance. Further, given the phase shifts between heating cooling (drought/enhanced rainfall) we consider the possibility that the tropical land masses contain significant climate memory or in conjunction with the ocean form joint memory of the climate system.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner