Figure 4. Results of Monte Carlo simulations assessing the significance of the link between (a) NH and (b) SH eruption age distribution and abrupt climate change over the interval 30–80 ka BP.

The root-mean-square (RMS) statistic is the summation of the distances between a set of eruption ages and the nearest abrupt climate change events (abrupt Greenland cooling events for NH eruptions, and warming events for SH eruptions) (see Supplementary Information). A set of eight NH eruption dates was randomly selected from a uniform distribution ten million times, and a set of five SH eruptions dates selected ten million times, and the RMS calculated. The blue frequency distributions illustrate the results of these simulations, and the vertical red lines illustrate the RMS of the actual eruption ages. The insets show the cumulative frequency distributions.