Determination of the Optimal Sampling Interval for Cyclostratigraphic Analysis by Using Sampling Theorem and Accumulation Rates

In recent years, cyclostratigraphy has been successfully applied to dating strata and recongnizing the possible astronomical forcing on major geological events. Sampling is one of the most important routines in cyclostratigraphic analysis to get the suitable geophysical or geochemical paleoclimate proxies. However, the workload will be significantly increased and random noises or other non-climatic noises will be introduced if the sampling frequency is too high; on the contrary, a lower sampling frequency may make it difficult to recognize Milankovitch signals in successions. In order to identify an optimal sampling interval, we used theoretic daily insolation data of time intervals of 80-100 Ma and two geological datasets to estimate each power spectra at three sampling intervals (high resolution, one quarter and half of one precession cycle), and then compared corresponding spectra analysis results. As a result, under the condition of satisfying the sampling theorem, sampling interval which equals to half of a precession cycle is the optimal sampling interval for cyclostratigraphic analysis. All Milankovitch signals can be identified and at the same time the workload is the least by using this optimal sampling interval. This interval should be determined according to the mean accumulation rate of the target successions during field sampling.