Reversal of three global atmospheric fields linking changes in SST anomalies in the Pacific, Atlantic and Indian oceans at tropical latitudes and midlatitudes

Abstract

Analysis of upper tropospheric relative humidity (UTH) variations from 1979 through 1992 shows the existence of a non-linear reversal in the tropical atmospheric anomalous circulation patterns characteristic of an El Nino (EN) phase. In addition to the tropics, both the original structures and their reversal affect midlatitudes. These changes begin a few months after the decaying of a strong EN episode and end one to two years after the peak in sea surface temperatures (SST) in the Pacific. This anomalous atmospheric circulation is matched by parallel reversals in cloud cover anomalies and absorbed solar radiation. At present, the limited time span covered by accurate satellite records prevents any conclusion regarding this reversal. An extension of coherent El Nino-Southern Oscillation (ENSO) signals in these correlation fields to higher latitudes is also reported. (By approximately between one to two years after the onset, the anomalous warming in the eastern equatorial Pacific had almost vanished, and the anomalies associated with the local Hadley cells and the Walker circulation revert towards a pre-El Nino stage). The coherence of these bimodal patterns might be responsible for certain climatic phenomena associated with ENSO in midlatitude regions. Though a relation between SST anomalies in the Pacific and the Indian and Atlantic oceans has already been postulated, this work shows how the return to near-normal conditions might yield a reversal in atmospheric patterns even in the absence of La Nina (LN) episodes. This in turn, might also produce climatic responses and feedbacks altering the net heat flux entering these remote regions, leading to droughts and floods. The results obtained strongly reinforce, for the most recent record, the hypothesis of a tropical atmospheric bridge connecting with the extratropics. In addition, it has been possible to isolate the ENSO signal from western Mediterranean (WM) SSTA, by means of advanced statistical techniques. The existence of anomalous atmospheric structures, that might link W Mediterranean SSTA and SSTA in the tropical North Atlantic (TNA), remains an open question, worthy of further investigation. Similarly, the role of ENSO in forcing this TNA region, and the subsequent atmospheric changes, will have to be further analysed with the aid of regional coupled models.