Reports: AC8
47231-AC8 Shapes, Scales and Spacing of Channel-belt Sand Bodies in Ancient and Experimental Avulsion-dominated Alluvial Basins
This project is focused on characterizing and understanding “avulsion clusters,” or groups of channel sandstone bodies separated from one another by intervals dominated by overbank deposition. Avulsion clusters are prominently exposed in the Ferris Formation of the Hanna Basin (south-central Wyoming) and may be common stratigraphic features in other alluvial basin fills. Sand body clusters have been shown to form in the absence of extrinsic forcing in physical and numerical experiment, although the extent to which autogenic avulsion clusters form in natural basins remains unclear.
The main effort of this year’s study was to statistically characterize sand body clustering in the Ferris Formation and use field data on channel properties to ascertain dominant controls on avulsion clustering within the interval. Most of this work comprised a PhD dissertation by Elizabeth Hajek at the University of Wyoming, completed in July 2009.
Using spatial point process statistics (specifically the “K function”), sand-body distributions in the Ferris Formation were compared to channel distributions from a physical experiment at the St. Anthony Falls Laboratory (University of Minnesota). The experiment (Delta Basin 2003) was run with constant boundary conditions in order to highlight intrinsic variability in the depositional system. Statistical evaluation of channel distributions within cross-sections of the experiment reveals spatial organization where channel deposits are regular (or anti-clustered) at short length scales (< 50 mm) and clustered at intermediate length scales (120-250 mm) within the basin. The Ferris Formation is random at short length scales (<100 m) and clustered over distances greater than ~ 100 m. This similarity between basins suggests that autogenic organization of channel avulsions over basin-filling timescales may have played a role in constructing the clustered stratigraphic pattern observed in the Ferris Formation.
In order to determine the potential for autogenic self-organization in this ancient deposit, the spatial patterns of auxiliary variables such as channel scale, maximum grain size, paleoflow depth, and paleocurrent direction were examined throughout the Ferris Formation in order to determine the factors that controlled channel avulsion and deposition within the basin. Stratigraphically there is evidence of a potential, albeit small, decrease in discharge up-section in the Ferris Formation (evidenced by decreasing bar-clinoform thicknesses), in addition to slightly shifting paleocurrent directions through time. These variables also show minor differences between some cluster groups. In contrast, sand-body grain size and scale do not vary stratigraphically or between clusters. Despite the minor differences in paleoflow depth and direction across the study area, the clustered pattern of stratigraphy remains the same. This suggests that in the face changing boundary conditions, autogenic avulsion clustering occurs and is the dominant control on sand-body architecture within the study area.
Further evaluation of channel variables within the Ferris Formation was conducted in order to provide baseline measures of variability within the interval. Descriptive sedimentology of a study area encompassing 120 channel sand bodies was completed. Channel sand-body architecture, variability in channel property measurements, and relationships between channel properties (including sand-body width, sand-body thickness, story thickness, story number, bar clinoform thickness, cross-bed thickness, paleocurrent orientation, and grain size) within individual channels were described. Additionally the effects of preservation and measurement error on documented variability were considered. Measured channel properties are scattered and do not show strong relationships with one another. One exception is story thickness, which is inversely related to the number of stories within sand bodies. Exposure quality does not strongly affect bar clinoform measurements, indicating that the range of observed variability is not an artifact of measurement error. Within multi-story sand bodies, a preponderance of fining-upward stories, fine-grained abandonment deposits preserved between stories, and large sedimentologic variability between successive stories are interpreted as evidence that multi-story sand bodies result from avulsion reoccupation.
Continued evaluation of sand-body distributions in experimental deposits is ongoing and should be completed during the remaining grant year. Channel deposits are being mapped in a cross section from another experiment conducted at the St. Anthony Falls Lab. This experiment experienced steady subsidence, mimicking a passive margin subsidence pattern, and several base-level fluctuations throughout the model run. We will compare channel distributions from different phases of the experiment to see if basin filling was affected by base-level cycles. Additionally, this experimental set up will be repeated in the winter of 2009/2010; this time, a foreland-basin subsidence pattern will be used. Channel deposits from that experiment will be mapped in cross section and statistically evaluated. Comparisons between experiments will highlight the role of subsidence changes on alluvial architecture.
Finally, a PhD student in statistics will be supported during the coming academic year. His role is push statistical techniques into the recognition and characterization of stacking patterns from well logs in various Rocky Mountain basins.
