H43H-01 INVITED
Of Magic Carpets, Rolling Snowballs, and Sleeping Dragons: an Energetics-based Classification for Hillslope/channel Interactions
Interactions between hillslopes and channels can include a diverse range of geophysical processes, including debris flows, landslides, water floods, and volcanic flows. Each has its own characteristic time-energy trajectory. In some cases the energy of an event increases as it propagates through a landscape, primarily through the addition of mass and momentum; examples of these"rolling snowball" include the initiation and runout phases of volcanic lahars, avalanches, and debris flows. In other cases, loss of both mass and momentum from a moving body or fluid causes the energy of an event to dissipate with distance, similar to the unwinding of a rug; examples of these "magic carpets" include the depositional phases of lahars, pyroclastic flows, lava flows, and debris flows. Both snowballs and carpets leave distinctive imprints or tracks on the landscape that reflect the resultant mass flux from hill slope to channel. The efficiency of this mass transfer depends on the width and slope of the receiving channel and the rheological properties of the transported material. At one extreme, the channel easily accommodates mass flux from the slope, sometimes accompanied by fractionation into constituent phases. At the other extreme, mass from the hill slope can inundate and block the channel; these "sleeping dragons" modulate subsequent mass transfer down channel by changing the channel profile and bed properties. They also have the potential to "wake up" suddenly as mass failure and/or outbreak floods. Hazard prediction requires that the time-energy trajectory of each type of event be assessed; here we suggest some first order controls.
H43H-02 INVITED
Hillslope processes and the delivery of sediment to the channel
There are several classic "chicken-egg" problems yet to be resolved in our pursuit of understanding how landscapes evolve. Here we focus on the problem of hillslope coupling with channels. As articulated by the conveners, a principal question is whether hillslopes are equal partners with channels in regulating landscape adjustment rates? Or, is there an imbalance such that either channel incision is driving hillslope evolution, or that sediment delivery from the surrounding hillslopes is setting channel incision? Naturally, there are many variables to control for, such as climate, tectonics and lithology, when trying to answer any of these questions. We try to constrain the numerous scenarios and special cases of landscape evolution by focusing on three classic characteristic profiles that can represent most landscapes. First, the equilibrium profile where, to the best of our knowledge, channels and hillslopes are eroding at roughly the same rate. Second, the inner-gorge profile, where we believe the channels are outpacing significantly the hillslopes. Lastly, the flattening landscape where the hillslopes are thought to be wearing down more rapidly than the channels are incising. Examples of each of these characteristic profiles are found under a wide range of climatic and tectonic settings, although they are more commonly used to typify specific conditions. We present data quantifying hillslope sediment delivery processes and rates from several field-based methodologies: cosmogenic nuclides for erosion rates, major and trace elements for chemical weathering, short-lived isotopes and optically stimulated luminescence analyses for sediment transport processes, and morphometry to define the characteristic form of the landscape. We use these data to propose an approach towards determining the dominance of hillslope or channel.
H43H-03
Sediment Transfer and Storage in Headwater Basins of the Oregon Coast Range: Transit Times from Radiocarbon-Dated Deposits
Many hillslopes in mountain landscapes such as the Oregon Coast Range (OCR) are coupled to channels through episodic delivery of sediment by debris flows to valley-bottom storage and gradual evacuation by fluvial processes. We characterized transit times through storage at two tributary confluences with the mainstems of Cedar and Golden Ridge Creeks in the central OCR with dense sampling of bank exposures: 68 radiocarbon age estimates from randomly assigned locations in the tributary and mainstem channel banks provided transit-time proxies. The inferred mean transit times were 1240 14C yrs for a debris fan tributary to Cedar Creek and 1510 14C yrs for fluvial terraces tributary to Golden Ridge Creek. Inferred transit-time distributions have double-exponential shapes for both tributary deposits and imply that younger deposits are preferentially evacuated: while most material moves through these storage sites rapidly (with characteristic timescales of 250 and 500 14C yrs for Cedar and Golden Ridge, respectively), a slower component is stored for millennia (with characteristic timescales of 2500 and 3300 14C yrs for Cedar and Golden Ridge, respectively). Juxtaposition of young and old bank exposures throughout both tributary deposits attests to repeated cycles of localized deposition and partial evacuation that produce steady-state sediment "reservoirs" over centuries and longer times. Flux estimates derived from the inferred mean transit times indicate that most (>66%) of the sediment yield of the Cedar tributary is stored in the fan for some time but only a small part (3%) is stored in the Golden Ridge tributary terraces. Whereas fluvial deposits like that at Golden Ridge Creek provide a relatively weak buffer between their contributing areas and larger downstream channels, debris fans like that at Cedar Creek provide the major coupling between hillslopes in such tributary basins dominated by debris flows and mainstem channels dominated by fluvial processes. The accommodation space necessary for debris fans and other debris-flow deposits therefore imposes a strong control on mainstem valley and channel widths.
H43H-04
Travel Distance in Landscape Evolution Models
Models for the evolution of tectonically active steeplands are commonly constrained by two limitations, the dynamic interactions with valley floor width and the balance between supply and flux limited removal. This paper discusses the implications of using sediment travel distance to provide a transition between supply- and flux- limited models. Where travel distance is long relative to the region of interest, removal is essentially supply limited; and where short removal is flux limited. For most sites therefore, the coarsest debris is flux limited whereas fines are supply limited. Once a significant fraction of the removal is flux limited, then models must also consider the supply of sediment from the hillsides which is generally dominated by mass movement, interacting with regolith forming processes dynamically linked to gradient. To make use of travel distance as a model variable, there is a need to define it operationally. Most formulations tend to treat it as an event-scale measure, describing total travel distance during an entire event, and most measurements of marked tracers inevitably adopt the same approach. Such measurements provide distributions of travel distance, related to grain size, which commonly correspond well with gamma distributions, having a modal distance that moves away from zero as the tracer source is exhausted. However the total movement of a grain is built up from a number of individual bed or suspended load movements, and it is at this scale that travel distance might be related to sediment transport mechanics. Some analogy is made with storm or daily rainfall amounts, that are also gamma distributed, but for which much more detailed information is available on intensity variations within a storm, to suggest possible distributions that underlie the total event travel distance.
H43H-05
A Hot Knife Through Ice-Cream: Earthflow Response to Channel Incision (Or Channel Response to Earthflows?), Eel River Canyon, California
Abundant glacier-like earthflow features are recognized as a primary erosional process in the highly erodable Franciscan Melange of the Eel River Basin, CA. Despite their prominence in this "melting ice-cream" topography, many questions regarding their effects on the long term sediment flux from this rapidly eroding basin remain unresolved. For example, does an earthflow's basal shear zone propagate vertically downwards with vertical river incision? What controls the upslope and lateral extent of individual earthflows? How does the erosive power of a river influence the rate of earthflow movement, or conversely do earthflow toe deposits regulate the rate of river incision? Here we present preliminary findings derived from study of 200km2 of lidar data (1m resolution) covering hillslopes adjacent to 30km of the Eel River. Lidar allows detailed analysis of the interaction between earthflows and the drainage network, and we document how inferred changes in local base level are propagated throughout adjacent hillslopes via earthflow movement. The most active earthflows (determined by field surveying and analysis of aerial photos rectified using lidar- generated digital topography) coincide with locally steep sections of channel, while downstream of the most active flows we frequently observe less-active or dormant earthflows. This observation supports the idea that the locations of the most active earthflows coincide with headward propagating knickpoints in the channel. The rate of earthflow movement appears to slow when an earthflow exhausts the upslope area of easily mobilized sediment. Earthflow toes can protrude directly into the channel, causing the channel to narrow and steepen, and even undercut the opposite bank. Large resistant boulders (>2m diameter) transported by the earthflow accumulate in the streambed and appear to both act as a check on further channel incision and earthflow movement. In contrast, areas adjacent to active earthflows exhibit smooth hillslopes, which show little or no evidence for recent instability. Such unfailed hillslopes preferentially occur near ridges or adjacent to strath terraces, and appear to be largely isolated from the effects of channel incision.
H43H-06
Progressive Landslides in Uplifted Volcanic Plateaus: Persistent Loci of Channel Perturbation
The semi-arid uplifted volcanic plateaus of the southern interior Columbia River basin contain over 300 large landslides or landslide complexes, ranging in area from several tenths of a km2 to several tens of km2. The distribution of these landslides is dominated by the outcropping of key stratigraphic contacts between coherent, volcanic cap rock atop weak sedimentary or volcaniclastic units in areas of >100 m local relief. The morphologies of many of these landslide complexes suggest a progressive mode of mass movement, with rubble-capped failure slices arrayed downslope at intervals of 10s to 100s of meters and deep tension cracks separating incipient failure blocks from the intact headscarp. Field evidence from the Owyhee River in southeastern Oregon indicates that individual landslide complexes can persist for millions of years. In one reach, for example, remnants of a 1.9 million year old intracanyon lava flow are inset against ancient landslide blocks. In the same location, cosmogenic isotope dating of boulders on a likely dam-burst flood deposit reveal a channel-blocking mass movement that may be as young as Holocene in age. The persistence of these landslide complexes has important implications for channel evolution, as it suggests that, in some environments, sediment supply may be chronically elevated at point sources. On the basis of GIS-based mapping of regional landslides and on field study of individual landslide complexes, we hypothesize that: 1) this sediment supply becomes increasingly dominated by fine-grained material as channels progressively incise into the weak units underlying coherent lava caps; and 2) the mass movements that impinge on the channels become correspondingly more earthflow-like. Loci of persistent landsliding are also subject to episodic variations in channel width due to physical constrictions caused by impinging failure masses. The discrete localization of large landslide complexes by particular stratigraphic and topographic characteristics and the potential for long- term channel perturbation driven by such complexes requires re-consideration of continuous representations of discharge, sediment supply, channel bed erodibility, and channel width in fluvial incision models.
H43H-07
The persistence of plateaus: Sediment-flux driven bedrock incision in the southern Sierra Nevada
Hillslope and channel evolution are tightly coupled in sediment-flux-driven models of bedrock channel incision. In this paper I explore the consequences of this coupling for the case of the plateau-dominated landscape of the southern Sierra Nevada, California. I use a sediment-flux-driven incision model to model the fluvial Cenozoic geomorphic evolution and surface uplift history of the range in response to different hypothetical uplift histories. Cosmogenic data for upland erosion and mainstem river incision rates allow the model parameters to be uniquely constrained. Numerical experiments using the sediment-flux-driven model suggest that the modern southern Sierra Nevada was constructed from a ~1.0 km pulse of range-wide surface uplift in the late Cretaceous (~60 Ma) and a ~0.5 km pulse in the late Miocene (~10 Ma). The persistent geomorphic response to late Cretaceous uplift in this model is the result of limited cutting tools supplied from the upland low-relief Boreal Plateau. This uplift history correctly predicts the modern topography of the range, including the approximate elevations and extents of the Chagoopa and Boreal Plateaux and their associated knickpoints. These results indicate that low-relief upland landscapes do not necessarily indicate late Cenozoic surface uplift. Finally, I illustrate some implications of the sediment-flux-driven model for understanding the persistent topography of ancient orogens.
H43H-08
Variable discharge controls on channel width and sinuosity in mixed bedrock-alluvial river channels: a combined field, remote-sensing, and modeling study
Rates of erosion in mountain river channels are the integrated effect of broad distributions of discharge rather than the result of a single, bankfull discharge. So, we can expect discharge variability, and the shape of the discharge PDF, to have a first-order effect on the morphodynamics of mixed bedrock-alluvial channels. Here we report the results of field surveys of Japanese and Taiwanese mountain rivers and analysis of corresponding gauging data, time-series satellite imagery, and SRTM topography. We show that: (1) bedrock channel width scales with both discharge mean and discharge variance; (2) wall sediment (by hillslope debris) and not just bed buffering must be considered; (3) relative discharge variability exerts on stronger effect on lateral channel mobility (meandering) than on typical channel width. To help explain such behavior we introduce a basic, dynamic model of bedrock channel geometry that employs a trapezoidal geometry, sediment buffering on both the channel bed (using queueing theory) and along the walls, semi-empirical channel hydraulics, and a schematic discharge PDF that maps (surprising well) the broad range of observed flood frequencies into a two-valued discrete distribution reflecting high-stage and extreme-stage flows. We solve for the model dynamics and derive a host of predicted relationships between discharge statistics and mixed bedrock-alluvial channel geometry at equilibrium and beyond. The model reproduces many of our empirical observations, in particular that bedrock channel meandering should accelerate with greater discharge variability. The model also suggests that increased bed sediment throughput drives channel widening, while increased fluxes of hillslope debris and transient storage along channel margins drives narrowing. We conclude there is a strong morphodynamic link between the shape of a mixed bedrock-alluvial channel and both the shape of its discharge distribution and its fluxes of hillslope and bed sediment. Our results point particularly to the need for more study of discharge variability and hillslope sediment buffering on erosion distributions in bedrock rivers. http://geomorph.ldeo.columbia.edu/grg