50 years of the turbidite paradigm (1950s-1990s): deep-water processes andfacies models - a critical perspective

Under the prevailing turbidite paradigm, the term 'turbidite' (i.e., deposi ts of turbidity currents with Newtonian rheology and turbulent state) is us ed very loosely and is commonly applied to deposits of debris flows with pl astic rheology and laminar state. For example, because 'high-density turbid ity currents' are defined on the basis of three different concepts (i.e., h ow density, grain size, and driving force), there are no consistent criteri a for recognition of their deposits. As a result, deep-water massive sands of debris-flow origin are routinely misinterpreted as high-density turbidit es. The concept of waxing flow as a type of turbidity current is problemati c because waxing flows are defined on the basis of velocity, not on fluid r heology and flow state. The waxing-flow concept allows inversely graded san ds to be misinterpreted as turbidites. Perhaps, the most problematic issue is the use of alluvial channel traction bed forms observed in flume experim ents as the analog for the five divisions of the Bouma Sequence (i.e., clas sic turbidites deposited from suspension). This is because flume experiment s were conducted under equilibrium flow conditions, whereas natural turbidi ty currents deposit sediment under disequilibrium waning flow conditions. T his and other problems of deep-water processes and facies models are addres sed in this paper from the author's personal perspective. Classification of sediment-gravity flows into Newtonian flows (e.g., turbidity currents) and plastic flows (e.g., debris flows), based on fluid rheology and flow state , is a meaningful and practical approach. Although popular deep-water facie s models are based on transport mechanisms, there are no standard criteria in the depositional record to reliably interpret transport mechanisms. Acco rding to existing turbidite-facies models, an ideal turbidite bed, which ha s normal grading, with gravel- to mud-size particles should contain a total of 16 divisions. However, no one has ever documented a complete turbidite bed with 16 divisions in modern or ancient deposits. Recognition of units d eposited by deep-water bottom currents (also referred to as contour current s) is difficult. Traction structures are good indicators of bottom-current reworking, but distinguishing deposits of bottom currents from deposits of overbanking turbidity currents is difficult even though it has important im plications for developing depositional models for hydrocarbon exploration a nd production. I consider sandy debris flows to be the dominant process res ponsible for transporting and depositing sands in the deep sea. Experiments on sandy debris flows suggest that low clay content (as little as 1%) is s ufficient to provide the strength necessary for sandy debris flows. Deposit s of experimental sandy debris hows are characterized by massive sand, shar p upper contacts, floating clasts, inverse grading, normal grading with cla sts, and water-escape structures. As a counterpart to turbidite-dominated f an models suited for basinal settings, a slope model is proposed that is a debris-Bow dominated setting with both non-channelized and channelized syst ems. Contrary to popular belief, deposits of sandy debris flows can be thic k, areally extensive, clean (i.e., mud poor), and excellent reservoirs. Hig h-frequency flows tend to develop amalgamated debris-flow deposits with lat eral connectivity and sheet-like geometry. Submarine-fan models with turbid ite channels and lobes have controlled our thinking for nearly 35 years, bu t I consider that these models are obsolete. The suprafan lobe concept was influential in both sedimentologic and sequea ce-stratigraphic circles because it provided a basis for constructing a gen eral fan model and for linking mounded seismic facies with sheet-like turbi dite sandstones. However, this concept recently was abandoned by its propon ent, which has left the popular sequence-stratigraphic fan models with a sh aky foundation. A paradigm shift is in order in the 21st century. This shif t should involve the realization that thick deep-water massive sands are de posits of debris flows, not 'high-density turbidites'. However, there are n o standard vertical facies models that can be applied universally for eithe r turbidites, contourites, or sandy debris flows. Science is a journey, whe reas facies models terminate that journey and become the final destination. (C) 2000 Elsevier Science Ltd. All rights reserved.