The lithofacies organization of fluvial channel deposits: A meta-analysis of modern rivers

Environmental interpretations of subsurface fluvial successions are commonly based on facies observations from core and are often attempted by generalist geologists by reference to classic fades models. However, for fluvial channel deposits, the value of observations on lithofacies proportions for interpretations of depositional environment has yet to be assessed quantitatively. Here, a test is presented that is based on a comparative study of facies data from 77 reaches of 46 modern rivers. The analysis is undertaken on datasets from published case studies stored in a sedimentological database, with consideration of causes for observational bias, and with particular attention paid to sandy lithofacies. The observed variability in the proportion of fades assemblages in the channel deposits of sandy river systems is quantified for classes of environments categorized according to channel pattern (braided, low sinuosity, meandering), climatic setting (arid to perhumid), and discharge regime (ephemeral to perennial). By capturing the variability in facies organization within fluvial systems of certain types, these outputs serve as facies models that provide a measure of uncertainty to sedimentological interpretations. Concurrently, the statistical analysis presented enables a test of the significance of relationships between the relative proportions of channel lithofacies and parameters that either represent controlling factors (e.g., water-discharge characteristics) or covariates (e.g., channel pattern). For classes of river systems grouped by channel pattern, climate, and discharge regime, emerging features of fades organization can be identified. Statistically, it is observed that relationships exist (i) between channel pattern and the frequency of the preserved expression of bedforms, and (ii) between controls on river hydrology (climate, discharge regime and seasonal variability) and the record of upper and lower flow-regime conditions. Thus, the results corroborate existing qualitative facies models in some respects. However, observations of the relative dominance of fades in channel deposits demonstrate limited value for interpretations or predictions in subsurface or outcrop studies, as variability within each type of depositional system is significant. Corehole data of fluvial channel deposits may be commonly overinterpreted. (C) 2019 Elsevier B.V. All rights reserved.