This study investigates the connection between hydraulic and sediment transport processes and channel and lobe geometry in supercritical submarine fans. The geometry of lobate deposits is typically explored as an interrelationship between different geometric measures, e.g., length versus width or thickness versus area, or based on confinement. However, the applicability of these relationships can be unclear due to their multi-scale nature and a lack of understanding of what processes control fan organization at different scales. Here we use a set of laboratory experiments and high-resolution seismic data from the Golo fan to distinguish three different hierarchical scales of nested lobate bodies with different characteristics and stacking (i.e., lobe element, lobe, and lobe complex) and compare characteristics of these mesoscale elements to equations describing 1D hydraulics along supercritical distributary channels. In particular, we are interested in defining the synoptic process and element in steep bedload-dominated supercritical fans that can most accurately be described as a function of hydraulic variables in the distributary channels. Based on the importance of hydraulic jumps in lobe-element formation in the experiments, we propose that the densimetric Froude number of the distributary channel exerts control on the geometry of the lobe element deposits associated with an individual avulsion cycle. Specifically, we hypothesize that higher Froude numbers will lead to lobate deposits that have larger ratios of lobe-element thickness to channel depth compared to those produced at lower Froude numbers. Both the laboratory and field data support this hypothesis. However, the conclusions are dependent on the correct hierarchical linkage between elements in the field and in experiments. In the analysis, the maximum lobe-element thickness was found to be well approximated by the sequent depth of the supercritical flow in the distributary channel. Therefore, this approximation yields a prediction of the lobe-element thickness based only on the hydraulic properties of the distributary channel without the need for any calibration or regression coefficients in the prediction.