Solar Influence on North Atlantic Mean Sea Level Pressure

Abstract

A principal component analysis of a 101 year set of winter mean sea level pressure anomalies for the North Atlantic-European sector has revealed evidence of an 11 yr cycle. This cycle is apparently related to the long-term modulation of solar flare activity during the sunspot cycle.

The principal components represent independent atmospheric circulation anomaly patterns. The variation of the strength of these patterns with time is given by a series of coefficients.

The variance spectrum of the coefficients of the second principal component (PC2) contains statistically significant peaks at 11, 5 and 3-1/2 years, indicating the presence of a non-sinusoidal 11 yr fluctuation. The main feature of the spatial pattern of PC2 is an area of intensified cyclonic activity to the west of the British Isles. This pattern, identified in seasonal data, is consistent with the pattern of 24 hr 500 mb height change following solar flare events reported elsewhere in this volume. The cyclonic intensification is located beneath a region of divergence ahead of an anomalous upper air trough. It is proposed that the 11 yr fluctuation arises as the frequency of solar flares, and the atmospheric responses on the day-to-day time scale, changes with the sunspot cycle.

This compatibility between the short term solar-weather and long term solar-climate relationships suggests an approach which may avoid a problem frequently encountered in investigations of the latter. Although the data length may be sufficient to generate a hypothesis, there is rarely additional data available to test that hypothesis rigorously. The approach would be to base hypotheses not on analysis of the limited climatic data, but on the results of short term solar-weather investigations, i.e., to deduce their implications, or integrated effects, on climatic time scales.

The phase locking between the sunspot cycle and the 11 yr fluctuation in PC2 was reasonable during the period 1920 to date, but broke down about the turn of the century. Although no firm conclusions can be drawn from the limited data available, three possible explanations felt worthy of further investigation are advanced. First, sunspot numbers may not be a good indicator of the component of the solar output affecting the atmospheric circulation. Second, the volcanic dust loading of the stratosphere was minimal during the period 1920 to 1960. This may have resulted in a stronger solar control on the atmospheric circulation during that period; at other times, the volcanic control may mask the solar effects. Third, the amplitude of the sunspot cycle was far greater after 1920 than it was before.