Previous: QBO and surface air temperature response to solar activity Next: Quasibiennial oscillations in North-South solar asymmetry Up: Ext. Abst.

Instability of the relations

For individual meteorological stations, as seen from [3], much better correlations can be found in the west QBO phase between surface air temperatures and sunspot numbers than for zonal and moreover hemispheric averages. However, in some but not all of them, a change of the sign of the correlation is observed around 1970 (Fig.7).

Changes in the sign of the correlation between solar activity and meteorological elements have been reported by a number of authors (see for example [2] and the references therein) based on nonstratified data. This fact is one of the main arguments against the solar activity influence upon climate. We have made a compilation of the available published results reporting positive or negative correlations between solar activity and surface air temperature in different locations (Fig.8).

It can be seen that the cases of positive and negative correlations are fairly well grouped, with the sign of the correlation depending on the period studied and not on the location, and changing in consecutive secular cycles. This observation is confirmed from reconstructions of global, hemispheric and zonal temperatures (Fig.9): in the end of the 18^th century high temperatures are observed in solar minimum and low ones - at high solar activity, and about 1920-30 the situation changes to the opposite.

As the global reconstructions are relatively short, to study earlier epochs data from individual meteorological stations with long measurement records were used from the Global Historical Climatology Network Temperature Database (GHCN) of NCDC [16]. The time series for each station was divided into subseries yielding the best correlation with solar activity. In Fig.10 the relative number of the stations with statistically significant positive and negative correlation (along the positive and negative parts of the Y-axis, respectively) is shown together with the secular (or “Gleissberg”) solar activity cycle.

Fig.10 confirms that in the 18^th century, in the vast majority of the available stations the surface air temperature was positively correlated with solar activity in the 11-year solar cycle, this correlation changed to negative in the 19^th century and to positive again in the 20^th century. As pointed out in [2], for the interpretation of the changes of correlation, other solar parameters should be considered apart from the sunspot numbers. As such parameter changing in consecutive secular solar cycles has been identified the North-South solar activity asymmetry [17] defined as A=(S_N-S_S)/(S_N+S_S) where S_N and S_S stand for the given solar activity parameter (sunspot number, sunspot area, number of major flares, etc.) in the Northern and Southern hemisphere, respectively. Negative A seen in the 19^th century means more active Southern solar hemisphere and is associated with negative correlation between solar activity and surface air temperature in the 11-year cycle, and positive A in the 20^th century - more active Northern hemisphere and positive correlation. As seen in Fig.10, the change in the sign of the correlation does not occur simultaneously all over the globe. About 1970 the asymmetry becomes negative and part of the stations begin showing negative correlations. This is the period when a change in the sign of the correlation is observed also in part of the stations in data stratified according to the phase of the stratospheric winds QBO. It could therefore be supposed that the factor determining the way in which solar activity affects climate is the North-South solar activity asymmetry while the phase of the QBO facilitates or hinders such an influence.