Sediment controls on the transition from debris flow to fluvial channels in steep mountain ranges

Abstract

Steep channel networks commonly show a transition from constant-gradient colluvial channels associated with debris flow activity to concave-up fluvial channels downstream. The trade-off between debris flow and fluvial erosion in steep channels remains unclear, which obscures connections among topography, tectonics, and climate in steep landscapes. Here, we analyze steep debris-flow-prone channels across the western United States and observe: (1) similar maximum channel gradients across a range of catchment erosion rates and geologic settings; and (2) lengthening colluvial channels with coarsening sediment cover. Following this compilation, we hypothesize that steep channel gradients are controlled by two competing thresholds of motion for bed-sediment cover: bed failure by mass-wasting and fluvial entrainment. We use downstream patterns in discharge, channel geometry, and sediment size to calculate discharges needed to mobilize sediment cover by both mechanisms across channels in the San Gabriel Mountains (SGM) and northern San Jacinto Mountains (NSJM) in southern California. Across steep colluvial channels in both landscapes, decadal discharges are below fluvial entrainment thresholds but above mass-wasting entrainment thresholds for D 50 (median) sediment sizes, consistent with recent debris flows captured by repeat imagery. Colluvial channel gradient is similar despite > 3Â contrasts in surface sediment grain size. In concave-up fluvial channels downstream, decadal discharges exceed fluvial entrainment thresholds, and mass-wasting is not predicted on lower gradients. In both landscapes, fluvial channels steepen downstream compared to gradients needed to mobilize sediment cover, which we interpret to reflect downstream increases in sediment flux. Coarser sediment supply in the NSJM than the SGM increases fluvial entrainment thresholds, which increases total channel relief in the NSJM by (1) lengthening colluvial channels shaped by debris flows and (2) increasing fluvial channel gradients. Our compilation and downstream analysis show how differing sensitivity of fluvial and debris flow processes to sediment grain size impacts the relative relief of colluvial and fluvial regimes in headwater channel networks.

Debris‐Flow Process Controls on Steepland Morphology in the San Gabriel Mountains, California

“…We observed that catchments in the SGM with faster erosion rates exhibit systematically larger A df (Figures 2b and 3a), an observation consistent with that by Neely and DiBiase (2023) that also reflects the lengthening of steep headwater, debris-flow valleys noted by DiBiase et al (2012). Specifically, the relationship between erosion rate and A df is approximately linear ( 𝐴𝐴 𝐴𝐴df = 0.37𝐸𝐸 + 0.08 ; Figure 3a; Table 1), such that faster-eroding catchments have steepland valleys that extend farther downstream.…”

“…In the SGM, DiBiase et al (2012) and Neely and DiBiase (2023), respectively, noted that steepland valleys were longer and A df is larger in more rapidly eroding catchments than in more slowly eroding catchments. Neely and DiBiase (2023), however, implicated sediment grain size and the relative ability of debris flows and fluvial floods to transport sediment as being more important for setting the location of A df than uplift rate alone. Nonetheless, Equations 1 and 2 not only capture the topographic signature of a specific process but additionally provide a framework for a morphologic proxy that can be used to extract tectonic information from topography.…”

“…They observed that S 0 should increase with faster uplift rates in a less-than-linear fashion (exponent <1) for cases where fluvial incision is minimal, which may occur at the uppermost reaches of steepland valleys. Although the nature of the relationship between uplift rate and S 0 in natural landscapes is generally unclear, observations from Neely and DiBiase (2023) indicate that S 0 may exhibit a positive, roughly linear correlation with erosion rate at modest erosion rates and potentially reach a threshold S 0 value at more rapid erosion rates. Because landscape evolution models should be able to reproduce observed relationships between key forcing mechanisms (e.g., uplift) and landscape morphology, further quantifying the relationship between A df , S 0 , and erosion rate could also help with the formulation and testing of a debris-flow GTL.…”

“…Some works have found that it is possible to produce longitudinal profiles with slope‐invariant steepland valley bottoms by using modified versions of fluvial GTLs such as the stream power incision model that incorporates episodic fluvial floods (e.g., DiBiase & Whipple, 2011; Lague, 2014; Lague & Davy, 2003), by placing steep valleys under the control of hillslope processes (e.g., DiBiase et al., 2012; Ouimet et al., 2009), or including stochastic bedrock landsliding in landscape evolution models (e.g., Campforts et al., 2022). Other processes that may result in deviations from power‐law slope‐area scaling include lower entrainment thresholds required to transport sediment at steep slopes and low drainage areas (e.g., Lamb et al., 2008; Prancevic et al., 2014; Recking, 2009), rapid weathering on steep slopes as topographic stresses produce bedrock fractures (e.g., Li & Moon, 2021; Moon et al., 2017; Neely & DiBiase, 2020; St. Clair et al., 2015), and downstream changes in the width of geomorphically effective flows, including water‐dominated floods and debris flows (Alessio et al., 2021; Neely & DiBiase, 2023; Yanites, 2018). In contrast, since Stock and Dietrich (2006), there has been little progress explicating the role of debris flows as geomorphic agents, and basic questions remain about the combined effects of fluvial and debris‐flow erosion on the form of steepland valley bottoms and channels and the capability of debris‐flow GTLs within landscape evolution models to reproduce this curved slope‐area signature.…”

“…In the SGM, DiBiase et al. (2012) and Neely and DiBiase (2023), respectively, noted that steepland valleys were longer and A df is larger in more rapidly eroding catchments than in more slowly eroding catchments. Neely and DiBiase (2023), however, implicated sediment grain size and the relative ability of debris flows and fluvial floods to transport sediment as being more important for setting the location of A df than uplift rate alone.…”