BERYLLIUM-7 SPECIFIC ACTIVITY IN SURFACE AIR AND ITS CORRELATION WITH METEOROLOGICAL VARIABLES, SOLAR ZENITH ANGLE, AND NUMBER OF SUNSPOTS

This analysis presents linear correlations of beryllium-7 specific activity in surface air with several variables measured in Helsinki, Finland, over 1987-2011. Measurements of beryllium-7 are linearly correlated with the following 10 variables: tropopause height; mean, minimum and maximum temperature; precipitation; atmospheric pressure; potential vorticity at 300 hPa and 200 hPa; solar zenith angle; and the number of sunspots. A time lag is also introduced into the calculations of Pearson’s linear correlation coefficients, i.e. the beryllium-7 specific activity measured on a day is correlated with a value of a given variable measured on some preceding day. In particular, the time lag is allowed to vary between 0 and 7 days for tropopause height, meteorological parameters and potential vorticity, and between 0 and 31 days for solar zenith angle and number of sunspots. In this way, we obtain sets of correlation coefficients for each pair of the analysed variables, in which we look for the maximum correlation coefficient as well as for the time lag on which this value is reached. In addition to investigating all of the available measurements, the time series are further split into seasonal subsets. Weak to moderate maximum correlation coefficients, ranging between 0.29 and 0.44, are found for the beryllium-7 specific activity with the tropopause height, temperatures and atmospheric pressure. These correlations are reached with a time lag of 0–2 days. Seasonal correlation coefficients show that the strongest linear relationships are obtained in summer. Similarly, a moderate correlation (0.48) with the solar zenith angle is reached with a time lag of 0 days, but the seasonal correlation coefficients in this case show a weaker relationship during different seasons. On the other hand, our analysis suggests that the correlations of beryllium-7 with precipitation, potential vorticity and the number of sunspots are not statistically significant.