A statistical test for correlation between crater formation rate and mass extinctions

Previous works have analysed the stationary periodicities of terrestrial crater formation rate and mass extinctions, and have compared the periods to discuss their causality. However, the coincidence of the periods does not necessarily mean that the two are correlated. Thus, in order to estimate the correlation directly, we test the null hypothesis that extinctions occurred independently of peaks in the crater formation rate. This hypothesis is rejected with a low significance level (3.0-7.1 per cent), even if the uncertainty in dating the craters is taken into account. Therefore, the rhythm of the mass extinctions with the time-scale of 30 Myr is highly correlated with that of the giant impacts. present the data regarding mass extinc tions and the formation ages of terrestrial craters. For the