Dallas H. Abbott and Ann E. Isley

Extraterrestrial influences on mantle plume activity

Earth and Planetary Science Letters(December 2002), 205(1-2):53-62

Index Terms/Descriptors: Earth; igneous rocks; impact craters; impact features; impacts; interplanetary comparison; komatiite; mantle; mantle plumes; mass extinctions; Moon; statistical analysis; time series analysis; ultramafic composition; volcanic rocks


We use time series analysis to compare the impact histories of the Earth and Moon with the record of mantle plume activity. We use events with errors in their ages of < or =150 Ma. The terrestrial and lunar impact records, when smoothed at a 45-Ma interval, correlate at a 97% confidence level. This high confidence level suggests that we have an adequate sampling of most of the major impact events on the Earth. We then test the idea that existing mantle plumes may be strengthened by impacts. When smoothed at a 45-Ma interval, strong plumes correlate with the terrestrial impact record at better than a 99% confidence level. No time lag is discernible between the data sets, which is expected given their present error level. When the time series are smoothed at a 30-Ma interval, there are 10 major peaks in impact activity. Nine out of ten of these peaks have a counterpart in either or both of the strong mantle plume or the mantle plume time series. As a result, the strong mantle plume and the impact time series correlate at the 97% confidence level. The mantle plume and the impact time series correlate at the 90% confidence level. Finally, the Deccan plume showed greatly increased activity immediately after the Chixculub impact. The results of our analysis suggest that large meteorite and cometary impacts may well increase the amount of volcanism from already active mantle plumes.