Quantifying trajectories of geomorphic river recovery through analysis of assemblages of geomorphic units: Aiding detection to inform river management

River recovery is defined as the trajectory of change that a reach takes towards an improved condition. In geomorphic terms, this includes improvement in both the physical structure and function of a river. While there are numerous case studies that track river recovery at the reach scale, there is a lack of work that characterises the patterns of recovery across diverse river types. Here we use ergodic reasoning to quantitatively analyse changes in the assemblage of geomorphic units (GUs) that occur for rivers that are known to be at different stages of geomorphic recovery. We aim to understand how the physical structure of different river types change as recovery occurs. In this study, we adopted a semi-automated method to map GUs using Open Access LiDAR and Sentinel remote sensing imagery. We analyse the assemblage of GUs for 78 river sections that span eight river styles, three valley settings and two bed material textures – sand and gravel. We find that the patterns of river recovery vary for different river styles. Confined and laterally unconfined rivers exhibit linear and non-linear increases in richness, abundance, evenness, and diversity of GUs during recovery. Partly confined rivers show more variable trends for these measures, and channelised fill rivers show decreased diversity. During recovery, sand bed rivers are more susceptible to adjustment than gravel bed rivers. The presence of benches and islands indicates that recovery is underway across most river types. A statistically significant increase in abundance and area of benches and pools, and a decrease in abundance and area of floodplain steps can also be used to indicate that recovery is underway. Additionally, in early stages of recovery, adjustments take place at the unit level. Then as recovery progresses, adjustments tend to occur at the sub-unit level as evidenced by bank-attached bars becoming more compound in structure. Determining the indicator GUs and changes in assemblages to ‘look out for’ while in the field or on remote sensing images can aid the detection and analysis of river recovery, providing invaluable insight for process-based and recovery-enhancement approaches to river management.