Fluvial floodplain deposits represent an important component of the continental stratigraphic record, but less attention has been paid to these fine-grained successions in comparison to coarser-grained in-channel deposits. This work seeks to understand how floodplain successions are constructed and preserved in the stratigraphic record. This is achieved through a comparative study of ancient preserved overbank successions (Jurassic Salt Wash Member, Morrison Formation; Cretaceous Castlegate Sandstone and Neslen Formation, Mesaverde Group – Utah and Colorado, USA) with a range of analogous modern overbank systems (e.g. Okavango River, Zimbabwe; Lli River, Kazakhstan; Paraguay River, Argentina; Saskatchewan River, Saskatchewan; Betsiboka Madagascar; Peace River, Alberta). These overbank systems have developed in response to different climatic regimes, in proximity to major parent channel bodies with different morphological arrangements, scales and styles of lateral migration and accretion, in basins subject to various subsidence regimes, and under different degrees of marine influence. The exhumed systems permit the complex juxtaposition of multiple flood deposits that form crevasse-splay elements, levees and flood-basin fines, to be constrained.
In overbank areas adjacent to major fluvial channels, crevasse-splay development drives floodplain aggradation. Constituent splay components are arranged at various spatial and temporal scales: (i) lithofacies arising from accumulation via fundamental flow processes; (ii) individual event beds (~1 m thick) comprising one or more lithofacies that thin and fine away from parent channels (single flood event timescale); (iii) A crevasse-splay comprising genetically related cosets of strata up to 3 m thick that record the initiation, growth, and abandonment of the splay (101 to 102 year timescale); (iv) splay complexes comprising genetically related splay bodies that emanate from more than one breakout point along a river reach, typically due to common mechanisms, (102 to 103 year timescale) (v) splay successions comprising a group of splay complexes that are not generated from the same breakout point related but which accumulated in a single flood basin (103 to 104 year timescales).
Data are stored in a fluvial architectural database, FAKTS. Results of this comparative study of both ancient preserved successions and modern systems has enabled the construction of quantitative facies models that serve to link the sedimentology and preserved stratigraphic architecture of crevasse-splays to external morphological form. Database output can be used to test the role of both allogenic factors, such as climate and basin type, and autogenic factors, such as parent channel form and intrinsic evolutionary behaviour, in controlling splay construction, accumulation and preservation.
