Geological models are important in aiding our interpretation of the rock record, particularly where outcrops, or subsurface data, is sparse. Conceptual models have been built and published for distributive fluvial systems (e.g. fluvial fans) where a predictive downstream and temporal changes in fluvial characteristics (e.g. channel body size) are present. However, few studies have assessed how much variability is observed across such systems and therefore should be present within our predictive models. This talk will present work that assesses how much variability can be present within facies models. Focusing on recently collected data from the Huesca distributive fluvial system (Spain) with comparison to the well-documented Salt Wash distributive fluvial system (SW USA).