Spring temperature variability relative to the North Atlantic Oscillation and sunspots — A correlation analysis with a Monte Carlo implementation

A reconstruction of past spring temperatures was analysed and compared to long-term records of the North Atlantic Oscillation (NAO) and sunspots. This palaeoclimate reconstruction, built previously using multiple proxy evidence of historical and natural sources from southwest Finland, explained approx. 70% of the instrumental temperature variance over a spring season (February–June) and covered the time period from 1750 to the present. Correlations between the NAO and sunspots appear markedly high and, based on Monte Carlo tests, statistically significant (p < 0.01) since around 1925. Correlation between sunspots and temperatures appeared notably high over the past 50 years, but this association could not be confirmed by the significance test. Correlations between the NAO and temperatures were high and statistically significant. However, the pre-1860s NAO–temperature correlations were lower than the correlations after that date. Previous studies have emphasized the possibility of enhanced solar forcing on NAO variability over the past 30–40 years broadly operative on decadal scales (with so far unresolved explanatory mechanisms). Our results correlate with this view to a statistically significant extent (p < 0.01) in the context of the past two centuries of regional climate variability. The NAO–temperature correlations were clearly stronger than the correlations between sunspots and temperatures. Moreover, the correlations were stronger between decadally filtered records. Consequently, the potential solar forcing on regional temperatures may have operated on decadal scales and been augmented by the NAO–temperature association.