The DO-climate events are probably noise induced: statistical investigation of the claimed 1470 years cycle

The significance of the apparent 1470 years cycle in the recurrence of the Dansgaard-Oeschger ( DO) events, observed in the Greenland ice cores, is debated. Here we present statistical significance tests of this periodicity. The detection of a periodicity relies strongly on the accuracy of the dating of the DO events. Here we use both the new NGRIP GICC05 time scale based on multi-parameter annual layer counting and the GISP2 time scale where the periodicity is most pronounced. For the NGRIP dating the recurrence times are indistinguishable from a random occurrence. This is also the case for the GISP2 dating, except in the case where the DO9 event is omitted from the record. year cycle is most pronounced and the dating highly reliable (Svensson et al., 2006). The preliminary extension of GICC05 shows substantial differences to the GISP2 dating before 40 kyr BP. This may indicate that the GISP2 record prior to 40 kyr BP becomes increasingly less trustworthy for detecting a periodic signal. Even though we perform our analysis for both the GISP2 and the NGRIP dating we will emphasize that there are good reasons to believe that the new multi-parameter annual layer counting applied to the NGRIP core is the most reliable and accurate of the two (Svensson et al., 2006).