Large-scale SST variability in the midlatitudes and in the tropical Atlantic

Research output: Book/ReportOther ReportResearchpeer-review


In this these the SST variability of the northern midlatitudes and the tropical Atlantic have been analysed. The analysis has been based on the comparison of the observations with a hierarchy of different coupled simulations. The analysis of the midlatitude SST variability has shown that the large-scale features of the SST variability cannot be simulated by a fixed depth mixed layer ocean model and that the spectral distribution of the SST is significantly different for an AR-1 process on time scales from seasons to decades. The processes that are important for these differences are the seasonal variability of the mixed layer depth, the wind induced mixing, which entrains water from the sub-mixed layer ocean, and the heat exchange between the mixed layer and the sub-mixed layer ocean. The observed increase in the SST variance from the interannual to the decadal time scale is due to the heat exchange between the sub-mixed layer ocean and the mixed layer and not, as in the MIX50 simulation, merely an effect of the integration of atmospheric noise. All these processes can be simulated by the local air-sea interactions in the dynamical ocean mixed layer MIX(dynamic). The analysis of the seasonal predictability of the SST in the MIX(dynamic) simulation indicates that the knowledge of the actual mixed layer depth is important to predict the SST development in summer and fall. In the analysis of the tropical Atlantic SST variability, it was found that the two dominant SST patterns of the observed SST and in all analysed CGCMs are centred in the northern and in the southern trade wind zones, whereas the correlation between the two patterns is not significantly different from zero. An interhemispheric dipole, or stated differently, an anti-correlation of the SSTs in the northern and southern trade wind zones, which could be important for rainfall anomalies in e.g. north-east Brazil, does therefore not exist. I conclude that the often cited dipole pattern is an artifact of the EOF analysis technique used. The fact that the simple slab ocean model produces the same pattern, indicates that the SST anomalies are forced by the atmosphere consistent with the Null hypothesis of SST variability. In the final chapter of this work I have introduced a new technique to study the response of the atmosphere to a given SST pattern in a coupled simulation. In this new technique the SST anomaly patterns or historical SST time series is introduced by an additional heat flux into the seasonal mixed layer ocean model. The comparison of the atmospheric response in the coupled simulation with the usual AMIP-type simulation has shown that the response in the midlatitudes can be significantly different and that the response of the atmosphere is very sensible to the structure of the given SST anomaly pattern. In general the new technique seems to be a good tool to study the atmospheric response to SST anomaly pattern in the midlatitudes and instead of introducing a fixed SST pattern or a given historical SST time series, the mixed layer simulation offers many other possibilities.

Original languageEnglish
Number of pages119
Publication statusPublished - 1 Jan 2000
Externally publishedYes

Publication series

NameExamensarbeit - Max-Planck-Institut fur Meteorologie
ISSN (Print)0938-5177

Cite this