Scatter in the plots of Dst(min) versus Bz(min)

In the plots of −Dst(min) versus −Bz(min) in cycle 23, in restricted ranges of −Dst(min), there was lot of scatter and the correlations were low. A bivariate analysis using geomagnetic –Dst(min) and auroral AE index did not improve the situation. Substitution of −Dst(min) by the ram-pressure-corrected true ring current Dst*(min) improved the correlations only slightly. Using the product V Bz instead of Bz(min) as a proxy for magnetospheric energy input did not improve the correlations, even for the 12 severe storm events in cycle 23. A multivariate analysis using −Bz(min) versus −Dst(min), with AU, AL and auroral particle precipitation index POES as additional indices improved the correlations and reduced the scatter. It was noticed, however, that for these severe storms, a major part of the solar wind input energy (∼85%) was still going to the Dst ring current, but this had an uncertainty of ∼15%. So in any severe storm event, the ring current could receive anything from 70% to 100% of the solar wind input energy.