Annualmeeting:2017 CSDMS meeting-108

From CSDMS






Browse  abstracts


Evaluating order vs. disorder in fluvial system deposits: A statistical analysis of grain size and thickness trends within vertical successions of sediment packages in the Ganges-Brahmaputra-Meghna delta, Bangladesh

Ryan Sincavage, Vanderbilt University Nashville Tennessee, United States, United Kingdom. ryan.s.sincavage@vanderbilt.edu
Steven Goodbred, Vanderbilt University Nashville Tennessee, United States. steven.goodbred@vanderbilt.edu
Peter Burgess, University of Liverpool Liverpool , United Kingdom. Peter.Burgess@liverpool.ac.uk


[[Image:|300px|right|link=File:]]The propagation of environmental signals through the sediment routing system and their subsequent preservation or removal from the rock record is a central theme in current stratigraphic research. The identification of cyclicity and order in stratigraphic sequences with regard to vertical facies successions, thicknesses, and grain size trends is often used as indicator of preservation of non-random, extra-basinal signals (i.e. climate, tectonics, and base level). However, it is less clear to what extent the processes that alter these signals post-deposition (re-working, scour, and erosion) enhance or diminish cyclicity and order within preserved sediments. Furthermore, stratigraphic trends are often identified in subjective, qualitative terms and may be based more on a priori perception of order derived from depositional systems models than statistically robust trends inherent in the sediment archive. Here, we use a statistical metric to objectively evaluate order vs. disorder in the stratigraphic record in an attempt to identify the likelihood of a disordered (random) response to orderly (non-random) depositional processes. We utilize a quantitative geochemical and sedimentological dataset from the Ganges-Brahmaputra-Meghna delta (GMBD) to identify distinct fluvial sediment packages (defined as meter to 10s of meters thick sand packages similar in scale and character to modern bar forms) and statistical trends in their vertical successions across the delta. We begin by considering that the boundaries of these fining-upwards packages are defined by >50% increases in grain size from one sample to the next in a vertical succession (although other thresholds are evaluated as well). A runs metric “r” is then calculated by identifying streaks of increasing or decreasing sediment package thicknesses and volume weighted mean grain size. This metric is then compared to the output of a Monte Carlo simulation of 5000 synthetic boreholes created by random shuffles of the observed borehole data to determine the likelihood of a similar succession of sediment body thicknesses and grain size trends being generated by chance. Preliminary results indicate that the vast majority of observed thickness successions in the GBMD are statistically “disordered”, with regional variability correlated to discrete geomorphic provinces within the delta. Of note, sediment thickness trends from the main braidbelt exhibit the lowest probability of being generated by random chance, followed by the lower delta plain, and lastly by Sylhet basin, a semi-enclosed sub-basin in northeast Bangladesh that has experienced episodic occupation by the mainstem Brahmaputra River throughout the Holocene. Similar results (with some notable exceptions) are found within grain size runs analyses, with Sylhet basin exhibiting the least amount of order with regard to vertical changes in grain size. Previous studies have identified Sylhet basin as a site of rapid mass extraction, suggesting a possible inverse relationship between stratigraphic order and rates of sediment extraction in fluvial systems. These results lay the groundwork for future studies in the utility of simple statistical measures in identifying random vs. ordered successions of sediment packages as indicators of process-response relationships preserved in the stratigraphic record.