Reach-scale morphological channel classifications are underpinned by the theory that each channel type is related to an assemblage of reach- and catchment-scale hydrologic, topographic, and sediment supply drivers. However, the relative importance of each driver on reach morphology is unclear, as is the possibility that different driver assemblages yield the same reach morphology. Reach-scale classifications have never needed to be predicated on hydrology, yet hydrology controls discharge and thus sediment transport capacity. The scientific question is: do two or more regions with quantifiable differences in hydrologic setting end up with different reach-scale channel types, or do channel types transcend hydrologic setting because hydrologic setting is not a dominant control at the reach scale? This study answered this question by isolating hydrologic metrics as potential dominant controls of channel type. Three steps were applied in a large test basin with diverse hydrologic settings (Sacramento River, California) to: (1) create a reach-scale channel classification based on local site surveys, (2) categorize sites by flood magnitude, dimensionless flood magnitude, and annual hydrologic regime type, and (3) statistically analyze two hydrogeomorphic linkages. Statistical tests assessed the spatial distribution of channel types and the dependence of channel type morphological attributes by hydrologic setting. Results yielded ten channel types. Nearly all types existed across all hydrologic settings, which is perhaps a surprising development for hydrogeomorphology. Downstream hydraulic geometry relationships were statistically significant. In addition, cobble-dominated uniform streams showed a consistent inverse relationship between slope and dimensionless flood magnitude, an indication of dynamic equilibrium between transport capacity and sediment supply. However, most morphological attributes showed no sorting by hydrologic setting. This study suggests that median hydraulic geometry relations persist across basins and within channel types, but hydrologic influence on geomorphic variability is likely due to local influences rather than catchment-scale drivers.

Sustainable concepts of ecologically functional rivers challenge engineers, researchers, and planners. Advanced numerical modeling techniques produce nowadays high-precision terrain maps and spatially explicit hydrodynamic data that aid river design. Because of their complexity, however, ecomorphological processes can only be reproduced to a limited extent in numerical models. Intelligent post-processing of hydrodynamic numerical model results still enables ecological river engineering measures to be designed sustainably. We have embedded state-of-the-art concepts in novel algorithms to effectively plan self-maintaining habitat-enhancing design features, such as vegetation plantings or the artificial introduction of streamwood, with high physical stability. The algorithms apply a previously developed lifespan mapping technique and habitat suitability analysis to terraforming and bioengineering river design features. The results not only include analytical synopses, but also provide actively created, automatically generated project plans, which are optimized as a function of an efficiency metric that describes “costs per m² net gain in seasonal habitat area for target species”. To make the benefits of these novel algorithms available to a wide audience, we have implemented the codes in an open-source program called River Architect. In this contribution, we present the novel design concepts and algorithms as well as a case study of their application to a river restoration project on the Yuba River in California (USA). With River Architect, we ultimately created an objective, parameter-based, and automated framework for the design of vegetative river engineering features. In addition, we are able to define a framework for stable and ecologically viable terraforming features, but part of the planning of earthworks is still left to expert assessment. Thus, improving the algorithms to plan terraforming of permanent, self-sustaining, and eco-morphodynamic riverbed structures based on site-specific parameters is one of the future challenges.

Physical habitat losses for Pacific salmonids in California’s Central Valley motivate stream restoration. Considerable river morphodynamics affect the sustainability of habitat enhancing interventions. In addition, the presence of large dams in many river catchments causes low sediment supply. This study revises existing stream restoration techniques for their ecologically efficient and physically stable embedding in a 36-km testbed river. Ecological efficiency is evaluated in terms of a commonly used hydraulic habitat suitability index. Physical stability results from 2D hydrodynamic modelling of bed shear stress during steady flows of different flood frequencies. We differentiate between terraforming, stabilizing and maintaining stream restoration techniques, which constitute three feature layers. The first layer, terraforming, includes artificial terrain modifications such as grading or backwater creation to generate new habitat. These features require stabilization, which is provided by the second feature layer. The stabilization (layer two) is achieved by bioengineering such as placement of streamwood, angular boulders and vegetation plantings. The third feature layer has the purpose to maintain newly created habitat, e.g., through artificial gravel injections. We illustrate the application of the three-layer-approach at one major restoration site of the lower Yuba River using a self-written Python package. Ecohydraulic 2D modeling was applied to designs with incremental layer additions to evaluate newly created spawning habitat and feature sustainability. This procedure represents a pertinent way for stream restoration planning, which avoids non-sustainable habitat enhancement features and implements ecologically as well as physically sustainable features only.

Reservoir sedimentation management has become an important topic to large dams in the United States due to their historical design, current age, and increased environmental regulation. Less attention has been paid to small dams in remote mountains even though they are facing a more urgent sedimentation problem due to their relatively small storage. This study aimed to explore the relation between reservoir operations and sediment erosion in a small dam’s backwater zone to seek potential alternative solutions to flushing and excavation. Mindful timing and magnitude adjustment of water transfer, water diverted by tunnels, through a reservoir was hypothesized to strategically redistribute sediment erosion for sites with water transfer/diversion facilities in the main channel. It was found that sediment erosion was increased by over 100% by turning the water transfer to the maximum level which is 12 times higher than mean annual discharge. With stage drawdown, the increment of sediment erosion was further increased by over 50% compared with water transfer only scenarios. Upstream inflow with occurrence from 5% to 25% was found to be the optimal hydrologic condition for water transfer. These results indicated that water transfer is a promising strategy to redistribute deposited sediment downstream of it with appropriate stage drawdown.

Designing where to plant riparian vegetation is a component of many river projects. Several mechanistic models have been developed considering biological, soil, hydrological, and hydraulic requirements that influence riparian vegetation growth. However, many models are not spatial explicit and there remains high uncertainty as to where plantings will survive or die. This study sought to determine if a machine learning (ML) algorithm could be trained to accurately characterize the complex set of site attributes that promote survival, and do so exclusively using metrics derived from airborne LiDAR. Results could then be used to guide planting strategies. The selected testbed river was 34 km of alluvial, regulated, gravel/cobble river where planting projects are common and have high mortality. The lower Yuba River, California, USA was mapped at sub-meter resolution in 2017. Our approach has four steps. First, a set of 32,000 vegetation presence/absence observations were randomly selected from LiDAR-derived polygons of naturally occurring established vegetation. Second, the river was split into 75 training, validation and test areas. Third, a set of 17 LiDAR-derived topographic potential predictors were computed at 0.91-m (3-ft) resolution. Finally, a Random Forest machine learning model was trained to best predict vegetation presence. The model results in a riparian vegetation presence probability map and has a “Area Under the Curve” (AUC) of 0.77. As probability values are difficult to interpret, a forage ratio electivity index analysis was performed with statistical bootstrapping. Results show that points with probability values > 0.8 had ~ 8.5 times more riparian vegetation present than would be likely from random chance at the 95% confidence level. Microtopographic ‘vector ruggedness’ was identified as the main driver for vegetation presence, followed by Terrain Ruggedness Index and Roughness. In conclusion, a ML model can identify where riparian vegetation planting are most likely to succeed and guide design. Our results also suggest that more attention should be paid to creating rugged microtopography under plantings to help cuttings and seedlings establish deposition critical for nutrition.

Statistical classifications and machine-learning-based predictive models are increasingly used for environmental data analysis and management. There now exist numerous classifications on the same topic but applied to different regions or spatial scales, such as geomorphic classifications. However, no quantitative meta-analysis framework exists to compare and reconcile across multiple classifications. To fill this gap, we jointly characterize statistical classifications and predictions by combining information theory and machine learning in three novel ways by: (i) measuring the degree of discriminatory information underlying a statistical classification; (ii) estimating the stability of the learning process with tuning entropy; and (iii) leveraging the sequential coarse-graining of information inherent to deep neural networks but absent from traditional machine learning models. This framework is applied through a benchmark of 59 millions models on a unique example of a single statistical classification methodology applied to nine different regions of California, USA. Regional results show that random forest consistently outperforms deep neural networks. In addition, a correlation analysis between regional characteristics, the level of discriminatory information of each classification, and the performance in statistical learning explains variations in performance and reveals the decisive role of the spatial scale of classification outputs. Because such a spatial scale is itself linked to the common situation of limited field sampling, directly comparing findings from statistical classifications and associated predictions appears seldom justified. A more desirable avenue to compare findings lies in combining data underlying statistical approaches in an interpretable and justifiable environmental data science.

Does river topography have stage thresholds for maintaining fluvial landforms, and if so how can they be quantified? Geomorphic covariance structure analysis offers a novel, systematic framework for evaluating nested topographic patterns in river corridors. In this study, a threshold in mountain river stage was hypothesized to exist; above this stage landform structure is organized to be freely self-maintaining via flow convergence routing morphodynamics. A 13.2 km segment of the canyon-confined Yuba River, California, was studied using 2944 cross-sections. Geomorphic covariance structure analysis was carried out on a meter-resolution topographic model to test the hypothesis. A critical stage threshold governing flow convergence routing morphodynamics was evident in several metrics. Below this threshold, narrow/high “nozzle” and wide/low “oversized” landforms that are out-of-phase with flow convergence routing morphodynamics dominated (excluding “normal channel”), while above it wide/high “wide bar” and narrow/low “constricted pool” landforms consistent with the flow convergence mechanism were dominant. Three-level nesting of co-located base-bankfull-flood stage landforms was dictated by canyon confinement, with nozzle-nozzle-nozzle nesting as the top permutation, excluding normal channel. In conclusion, this study demonstrates a significantly different and highly effective approach to finding process-based fluvial thresholds that can complement pre-existing methods, such as estimating incipient sediment motion, to get at more powerful dynamics controlling fluvial landforms structure.