Proxies for solar EUV in cycle 23

Abstract

For daily values in a strong 27-day sequence, lasting for more than 8 months, correlations of solar EUV with solar <i>F</i>₁₀<i> </i>radio emission were, sometimes, found to be lower than those with solar Mg II core-to-wing index, sometimes higher, and<i> </i>sometimes almost the same. Correlations of EUV with the UV Si III and Lyman-α were found to be mostly higher than<i> </i>those with <i>F</i>₁₀<i> </i>and Mg II. In contrast to the correlations for long-term changes (0.95 or more), the maximum correlations for<i> </i>short-term changes (daily values) were only -0.80, implying an explained variance of only -65%, leaving -35% unexplained.<i> </i>Hence, estimates of daily EUV using <i>F</i>₁₀<i> </i>and Mg II as proxies could have uncertainties of several tens of percent.<i> </i>For intermediate and long-term changes, EUV was very highly correlated with <i>F</i>₁₀<i>. </i>Mg II and Lyman-α (and to some extent<i> </i>even with sunspots) during years of low and medium sunspot activity (1996-1999) of cycle 23. The Mg II had a slight superiority<i> </i>over <i>F</i>_10,<i> </i>but both could be considered as good proxies for EUV. However, for the sunspot maximum years 2000-2002, the EUV evolution differed from the evolutions of <i>F</i>₁₀and Mg II, so that these would be unsatisfactotry proxies for<i> </i>EUV.