Statistical analysis of turbidite cycles in submarine fan successions: Tests for short-term persistence

To statistically test and evaluate the significance of asymmetric upward th ickening and thinning trends and other cyclic patterns in turbidite success ions, 28 bed-by-bed sections with a wide coverage in geological time, tecto nic settings, facies characteristics, and depositional environments were me asured and described. First, 286 sandstone packets were selected from the 2 8 turbidite sections through statistically based segmentation. Then, these packets were examined by three powerful correlation tests (Kendall's, Spear man's, and Pearson's correlation tests) and four tests for randomness. Only 34 (11.9%) of the sandstone packets pass tests designed to identify asymme try at the 10% significance level, Monte Carlo simulation and calculation o f cumulative binomial probabilities indicate that the number of asymmetric cycles identified in the original set of sandstone packets is indistinguish able from the number expected to result from random processes. Eighty-six sandstone packets from five sections were tested for upward coar sening and fining trends, Forty two of these packets were from channel depo sits and the rest from lobes and basin-plain deposits, Only 25 (29%) of the 86 packets coarsen or fine upward. However, 33-50% of packets from coarse- grained channel fills show asymmetric grain-size trends, mostly as upward f ining cycles (15 of the 42 tested packets from channel deposits). These upw ard-fining trends in channel deposits are interpreted as the result of chan nel filling or the stacking of onlapping deposits at a channel mouth. Results from this study strongly suggest that asymmetric bed-thickness cycl es have essentially no statistical significance in turbidite successions, a nd therefore cannot provide a key criterion for identification of sub-envir onments in submarine fan systems. It follows that models for submarine fans based on the widely publicized hypothesis of common asymmetric cycles shou ld be reconsidered. Instead, other criteria such as specific facies charact eristics, large-scale geometry, and degree of sand bed clustering may provi de the best tools for discrimination of submarine fan sub-environments. In natural systems, the disordered vertical arrangement of thicker and thin ner beds likely results from irregular variations in flow volumes, concentr ations, and precise pathways of turbidity currents. Over the short time sca les represented by tens of meters of section, most submarine fans are built by aggradation, not progradation, so that facies shifts are controlled mos tly by channel switching and avulsion, and not by progressive basinward or landward advance or retreat of one sub environment over another. Aggradatio n does not tend to generate asymmetric trends in bed thickness. The same ar guments are believed to hold for grain-size trends in submarine-fan packets , except in the case of coarse-grained channel deposits, where upward finin g is conspicuous.