Brian S Burnham , Rhodri M. Jerrett, David Hodgetts, Stephen S. Flint

Stacked fluvial distributary channel deposits and palaeovalley fills can form major, multi-storey sand bodies with similar thicknesses, and with lateral extents often greater than a single exposure. Consequently, they can be difficult to tell apart from one another using outcrop data. This study addresses this problem by quantitatively analysing the architecture of five stacked fluvial distributary channel deposits and two palaeovalley fills from the Pennsylvanian Pikeville and Hyden formations of the central Appalachian Basin, USA. The a priori interpretation of the sand bodies as stacked distributary channels and palaeovalley fills is possible because a robust in-place coal seam correlation framework allows for the recognition of different basin-scale architectures for each type – aspect ratios <1000 and envelopes of fluvial and deltaic strata for stacked distributary channels, and aspect ratios >1000 and a regional basinward facies shift at the bases of palaeovalley fills. Sand body thickness, storey thickness, position and length of storey contacts within the sand body are similar in both types. However, they can be distinguished by different up-system to down-system changes in their respective architectures. Stacked distributary channel sand bodies thin down system, display a decrease in storey thickness, an increase in the mean position of storey heights in the sand body and a decrease in the length of storey contacts. These trends are the result of down-system decrease in channel size, and confinement associated with radially distributive fluvial systems. Palaeovalley fill sand bodies thicken down-system, display an increase in storey thickness, a decrease in the mean position of storey heights, and a decrease in the length of storey contacts. The increase in sand body and storey thickness are the result of down-system increases in original channel size, consistent with trunk axial fluvial systems fed by tributaries that predominate during valley-formation. The down-system increase in amalgamation reflects a down-system decrease in accommodation, from the higher subsidence rate active margin of the basin, and is therefore not necessarily characteristic of palaeovalley fill architectures in all basin settings. This study emphasises the requirement for detailed correlation work and quantitative analysis of external and internal architectures before the interpretation of sand bodies as stacked distributary channels or palaeovalley fills.