Relative elevation topographic surface modelling of a large alluvial river floodplain and applications for the study and management of Riparian landscapes

This paper presents a novel and useful spatial modelling technique to create a topographic surface that estimates a floodplain's elevation relative to the average low-flow water surface elevation of a river channel. This model was applied to a 121km study area of the middle Sacramento River, California, USA, where it was tested as a surrogate for observed water table depth and an observed 3.3 year recurrence interval flood inundation surface using independent data sets. The modelled relative elevation topographic surface correlated significantly (p < 0.005) to observed well water depths suggesting that the modelled surface reflected a reasonable approximation of vertical distance to the water table. Results from a flood inundation pattern analysis indicated an overall accuracy of 79% for correctly predicting inundated and non-inundated zones. The model was then used to measure relative channel bank height and the distribution of riparian plant communities to examine landscape ecological relationships.