H51C-1128 0800h
Are They Beach Cusps or Beach Horns? A Kauai Observation
Beach cusps on a southwest Kauai shoreline exhibit departures from the current schools of thought on cusp dynamics. Beach sediment is either carbonate sand or volcanic sand. The horns of the beach cusps are depositional features, reflecting the migration of sediment along the beach during the waning storm-wave (energy) phase and represent a building of the beach face. This phase may be comparable to the development of distinct sediment-transport bedforms in a unidirectional flow environment that develop at distinct ranges of Froude number. Heterogeneities such as groundwater-sapping patterns after storms, protruding bedrock, and variation of sediment size along the initial beach front (perhaps as a result of previous cusp development) provide the initial irregularities to initiate cusp morphology with evolution to a more rhythmic pattern because of the major feedback processes in the beach cusp environment. Although cusps may be stationary at some locations, transport of sediment along the shore is still occurring. The term beach horns, instead of beach cusps, should be used because the morphology is developed as a consequence of horn development. At the start of the field observation, long-period (LP) waves from south-southwest of Kauai dominated the shorter period Trade-Wind (TW) waves which are refracted around the island. The LP waves generated high-surf warnings. At this time, the cusps had an asymmetric pattern with the main backwash close to the horn located on the side of the cusp that represents the direction of longshore transport. Active erosion of the horn took place on the side of the cusp where more backwash occurred. The horn crests had an acute angle towards the approaching LP wave front. As the energy of the LP waves decreased, the TW waves became more dominant in the interference pattern of the two wave forms. The beach cusps reflected this change of interference in terms of symmetry as they became more symmetrical. Significant transport of sediment within the cusp feature was limited to high-tide periods. Troughs form offshore in front of the horns and sediment buildup occurs offshore of the cusps. Backwash in the cusp interferes with incoming waves, resulting in sediment deposition. In contrast, waves impinging on the horns had no interference from backwash and eroded the offshore troughs, depositing the sediment on the surface of the horn with additional sediment being carried into the cusps. Beach cusp processes were the same on the two types of beaches along this shoreline.
H51C-1129 0800h
Focused Weathering Control of Convex Waterfalls Along the Niobrara River in Cherry and Brown Counties, Nebraska
More than 200 waterfalls exist along the southern spring branch tributaries that feed an approximately twenty-five mile section of the Niobrara River, east of Valentine, Nebraska. Many of these waterfalls posses a convex shape in the horizontal plane and are buttressed. This morphology is controlled by focused, season-specific weathering along the escarpments adjacent to the waterfall face and a lack of stream erosion on the actual waterfall face. The waterfalls are composed of the Rosebud Formation, a poorly indurated siltstone that should be easily eroded by stream flow. The spring creeks are ineffective at significantly eroding the waterfall face mainly due to their relatively low discharge, three to five cubic feet per second, and low sediment load. The erosive power of the streams is further reduced at the site of the waterfall by the buttressed shape spreading the flow into a thin sheet. The buttressed shape of the waterfall develops in response to stress relief. The only areas of the waterfall face showing stream erosion and lack of diatom cover is where free falling water is impacting the waterfall face. Large, loose talus slopes at the base of the waterfall escarpments further support that the weathering processes operate at a faster rate than stream erosion. Observable groundwater seepage from the escarpments on either side of the waterfalls exposes the faces to season-specific weathering processes. The moisture content of the escarpments varies with exposure to sunlight and changes in air temperature. Cyclic differential expansion and contraction of clays and minerals as well as precipitation and hydration of salts operate on a daily and seasonal basis. These repeated stresses give the escarpments a flaky, shingle like appearance and can cause rapid deterioration of the escarpment. During the winter, the seeping groundwater and waterfall spray form large ice flows on either side of the waterfall face. Freeze-thaw processes operate on a seasonal and shorter cycle due to the variable regional weather. These weathering processes are focused on areas of the escarpment lacking stream flow. The stream protects the waterfall face by preventing cyclic moisture level changes during the summer and freeze-thaw cycles during the winter. With increasing convexity, the stream flow is diverted to the sides of the waterfall face. The waterfall face is then exposed to moisture level changes and freeze-thaw cycles due to lack of stream protection. The waterfall face experiences major failure and erodes upstream, whereby it is protected by stream flow again and begins to regain a convex shape.
H51C-1130 0800h
Tectonic and Climatic Controls of Alluvial fan Geometry
Alluvial fans are useful natural laboratories for the study of the tectonic, climatic, and lithologic boundary conditions that constrain landscape form. We evaluate the importance of these boundary conditions by building on the analysis advanced by Whipple and Trayler (1996, Basin Res., 8:351-366), and thus in terms of mass conservation and a simple stream-power erosion law applied to fan-catchment systems at topographic steady state. Our new analysis links fan area with rates of uplift and precipitation, area and relief of the eroding catchment supplying debris to the fan. The explicit relationships that emerge from this analysis provide a useful rationale for summary measures of fan-catchment geometries of previously-studied and environmentally wide-ranging systems in California, Argentina, Italy and Spain. In particular, we predict and observe relationships between the ratio of rock-uplift rate and precipitation rate and, respectively, the average relief of upland, source catchments and the ratio of fan area to the product of catchment area and relief.
H51C-1131 0800h
Changing Climate and Wind Patterns Revealed in Indiana's Fair Oaks Dunes
Fair Oak Dunes (FOD) cover over 1100 square miles in north-central Indiana. Careful study of dune morphology reveals three types of dunes in regards to their size. The first order forms are compound parabolic dunes that reach over five miles in length and have the apex of parabola pointing in a southwesterly direction. The spacing between these dunes is three to five miles. The second order dune ridges are compound parabolic dunes that range in size from one to three miles in length with spacing of about one mile between the ridges. Both, the second order and the third order dunes have the apex of parabola pointing in northeasterly direction, opposite of the first order dunes. The third order dune ridges are simple parabolic dunes that reach up to half mile in length and are 25 to 30 feet tall in western part to over 45 feet in the eastern part of the FOD. All dunes are fixed by lush vegetation. Preliminary grain size analyses indicate that north part of FOD has coarser sand (0.283 mm) than southern part (0.197 mm), while eastern part (0.271 mm) is coarser than the western part (0.223 mm). This grain size distribution is in accordance with initial interpretation of dune morphology. Strong northeasterly winds associated with anticyclone were prevalent in early dune formation about 14,000 years ago near the end of last glacial. The finest particles were blown south and southwest from the source area which was north and east of the present dunes. Cyclonic southwesterly winds become dominant in Holocene and caused a reworking of the original large dunes into smaller forms as well as removal of some of the finest particles back to the original source to the northeast. Limited vertical dune profiles indicate that below the 5 feet of bioturbated surface layer are alternating light layers (3 to 5 inch thick) and dark laminae (1-2 inches thick). Dark laminae consist of quartz grains with `hairy' surfaces covered with reddish iron oxides or clays. They contain twice as much silt fraction and are more indurated and resistant to weathering than light layers separating them. Are these alternating layers depositional or postdepositional structures? Depositional origin would indicate episodes of drier climate and strong winds creating light, coarser layers while dark laminae would be indication of wet climate and weak winds transporting only dust particles. Postdepositional origin would indicate formation of "dissipation structures" created by soil water translocation of fines during humid climate. Many of the questions about timing of original dunes and their later transformation as well as timing of dark laminae and light layers within the dunes will hopefully be answered by OSL dating.
H51C-1132 0800h
A Numerical Study of the Factors Controlling Valley Spacing in Landscapes
Landscapes often exhibit a characteristic spatial scale of dissection, which is most apparent as the spacing between adjacent valleys (or the intervening ridges). Simple physical and geometrical arguments imply that valley spacing could be determined by erosion thresholds and/or by the relative rates of advective (e.g., stream incision) and diffusive (e.g., soil creep) sediment transport processes. We solve a simplified, nonlinear advection-diffusion equation numerically to investigate the factors controlling valley spacing. Using dimensional analysis, we infer the form of a relationship between valley spacing, the rates of the advective and diffusive processes, and spatial and temporal boundary conditions. Numerical solutions of the advection-diffusion equation confirm the general form of this relationship. We show that a similar relationship should hold for more complicated expressions that are more realistic representations of sediment transport processes in nature. Our results also provide insight into the factors controlling local relief, the causes of transitions from rill-like drainage networks to branching networks, and the upper and lower bounds on the length-to-width ratios of first-order drainage basins.
H51C-1133 0800h
Terrestrial Sediment and Nutrient Discharge, and Their Potential Influence on Coral Reefs, Puerto Rico
Sediment and nutrient discharge to the insular shelf of Puerto Rico (18 degrees latitude), augmented by anthropogenic activity, is believed to have contributed to widespread degradation of coral reefs of Puerto Rico during the 20th century. Sediment deposition degrades coral reefs because it reduces the area of sea floor suitable for growth of new coral, diminishes the amount of light available for photosynthesis by symbiotic algae that live within individual coral animals, and in extreme cases, buries coral colonies. Land-use history and data from 30 water-discharge, 9 daily and 15 intermittent sediment-concentration, and 24 water-quality gaging stations were analyzed to investigate the timing and intensity of terrestrial sediment and nutrient discharge into coastal waters. Watersheds in Puerto Rico generally are small (10's to 100's of square km), channel gradients are steep, and stream valleys are deeply incised and narrow. Major storms are usually brief (<24 h) but intense such that the majority of the annual sediment discharge occurs in a few days. From 1960 through 2000 the highest mean daily discharge for a water year (October - September) accounted for 20 to 60 percent of the total annual sediment discharge. Major storms, with a return frequency of approximately a decade, were capable of discharging up to 30 times the median annual sediment-discharge volume. Prior to agricultural and industrial development, coastal waters are believed to have been relatively transparent, with strong currents and seasonal high-energy swells assisting corals in the removal of minor amounts of sediment deposited after storms. Land clearing and modification, first for agriculture and later for urban development, have increased sediment and nutrient influx to the coast during the 19th and 20th centuries. Although forest cover has increased to approximately 30 percent of the surface of Puerto Rico during the past 60 years, sediment eroded from hillslopes during the agricultural period is still being episodically transported from upland valleys to downstream floodplains and the coast. In response to better land management, the quality of water has improved significantly since the 1980s. Nitrogen and phosphorous concentrations in river waters are now well within regulatory limits, although current concentrations are as much as 10 times the estimated pre-settlement levels. Concentrations of pathogens also are improved but continue to be near or above regulatory limits. Unlike sediment discharge, which is episodic and intense, the discharge of river-borne nutrients and pathogens is a less intense but chronic stressor to coral reefs located near the mouths of rivers.
http://pr.water.usgs.gov/
H51C-1134 0800h
Location and Lateral Migration of Drainage Divides
We investigate the mechanisms driving the lateral migration of drainage divides on hillslopes experiencing dispersive-like soil creep due to changes in downcutting rates in the channels bordering these hillslopes. If two channels are downcutting at the same rate and are at the same elevation the hillslopes will be symmetric about the divide. This symmetry may be broken by a vertical offset in the elevation of the channels or by unequal rates of channel incision. The rate and distance of divide offset will depend on the dominant sediment transport mechanism on the hillslope. For hillslopes undergoing sediment transport that is linearly proportional to the slope, the divide offset distance is one quarter of the ratio of the vertical offset of the channels to the relief of the symmetric hillslope. If channels are downcutting at different rates, the speed of the divide migration will only depend on the ratio of the two downcutting rates and the density ratio, which is the ratio of the dry bulk density of the bedrock to the dry bulk density of the soil. The density ratio plays a fundamental role in determining the transient response of the divide if the bounding channels are lowering at the same time-averaged downcutting rate but whose instantaneous incision rates vary. Lower density ratios lead to faster transient responses of the divide to changes in channel downcutting rates. Other parameters that affect the transient response of the divide are the magnitude of transient differences in downcutting between the two channels, the time averaged incision rate, and a ratio of the relaxation time of the hillslope to the particle residence time on the hillslope. The soil depth profile reacts to transient changes in downcutting at a different rate than surface topography. Hillslopes experiencing transient channel downcutting may have surface topography that is symmetric about the divide but will at the same time have a soil depth profile that is asymmetric. The implications for long term adjustments of drainage basin morphology are considered.
H51C-1135 0800h
Valleys and Hillslopes: A Geomorphic Foundation for Landscape Ecology
Moisture-nutrient gradients have been found to be the most important environmental gradients determining the distribution and composition of plant communities. Landscapes on which plant communities exist are composed of valleys and ridgelines, with hillslopes in between them. Since water flow paths are directed down slopes, processes determining hillslope morphology and arrangement play an essential role in plant community organization and dynamics. Hillslope morphology, substrate characteristics and climate determine flow routing and water budgets along slopes. Wetness is a function of transmissivity, contributing area and slope gradient. Movement of nutrients along hillslopes generally follows wetness values, and is affected by soil type. Plant species have different tolerances to wetness and nutrients; hillslope length and slope angle determine the moisture-nutrient gradient, and in turn the shape of plant tolerance curves. Temporal scales required for significant topographic change along hillslopes may often be long compared to those for plant community dynamics. When considered in landscape ecology, hillslope shape and arrangement are thus often considered constants. Although landscape morphology may change over time and among different regions (with tectonic, geomorphic and climatic processes leaving their imprints on landscapes), an attempt has been made in the literature to put forth robust topographic scaling relations. This paper, using a series of examples, explores connections between landscape structure and plant communities. For example, Hack's law states that drainage basins become more elongate as area increases. This implies that basins should have approximately the same proportion of landscape in each hillslope position, suggesting some constancy in contributing area patterns for hillslopes in different-sized basins. Distributions of wetness values and plant population tolerance curves seem to confirm this for smaller basins. Hillslope length and steepness are related to drainage density and relative relief. Various studies have sought relations between drainage density and slope gradient; the latter is a determinant of wetness values. Studies have found both negative and positive correlations between drainage density and slope gradient. The nature of hillslope processes (e.g., overland flow vs. mass wasting dominated, or quickly eroding vs. slowly eroding landscapes) has been used to explain the correlation. It has also been suggested that the degree of channelization may be important in determining slope steepness. Plant species respond to steeper slopes by having narrower tolerance curves and less overlap with other species. This has important implications for biodiversity and plant community organization.
H51C-1136 0800h
Hanging Valleys in Fluvial Systems: A Failure of Stream Power and Implications for Landscape Evolution
We document the existence of hanging valleys in an unglaciated landscape in the eastern Central Range of Taiwan. Analyzing fluvial long profiles from over 140 channels, we establish a set of morphologic criteria defining a hanging valley, including: 1) channel gradient at the tributary mouth; 2) position of the oversteepened reach in relation to the tributary junction; and 3) height of the perched tributary lip above the trunk stream. Using these criteria, we suggest that 25 tributary basins in this fluvial system are effectively hanging valleys. Our conceptual model for the initiation of hanging valleys is consistent with a recently published mechanistic model of bedrock incision, which suggests that the highest transport stage flows are actually less efficient at eroding bedrock than moderate stages. As tributary mouths steepen and transport stage increases in response to an incisional pulse in the mainstem, channel gradients may pass a threshold value beyond which erosional efficiency is hindered, giving rise to a mismatch between trunk and tributary erosion rates. This mismatch is naturally expected at tributary junctions, where a step-function decrease in drainage area also leads to sharp contrasts in water and sediment discharge between trunk and tributary basins. The presence of hanging valleys in fluvial landscapes suggests that our most simplified parameterizations of bedrock erosion -- which typically assume a monotonic positive correlation between channel gradient and incision rate -- may be applicable only to a range of moderate channel gradients. More importantly, the presence of these features may have important implications for landscape evolution, since the catchments above these oversteepened reaches may be insulated for some time from incisional signals propagating through the channel network, increasing landscape response time beyond that expected based on the stream power incision model. The results of this study underscore the need for a more complete understanding of bedrock erosion processes and their incorporation into landscape evolution models. Finally, the presence of hanging valleys in Taiwan indicates significant perturbation away from steady-state conditions in the eastern Central Range.
H51C-1137 0800h
The pattern of coral reef development at two sites in the Seychelles
Extensive drilling and dating of modern reef complexes worldwide have shown them to be just thin veneers of Holocene material deposited on older Pleistocene limestone. Where the Pleistocene substrate has been imaged seismically it is often characterised by rough karst topography that is thought to be the result of exposure to subaerial weathering during sea-level low stands. The importance of this antecedent karst topography relative to organic growth processes in controlling present day reef morphology has been much debated. We present high-resolution side-scan backscatter and 2D bathymetric survey results from two sites in the Seychelles (western Indian Ocean). These sites are relatively protected and reef morphology is not modified by tropical storms. The sonogram mosaics reveal areas of uniform backscatter intensity, together with regions of coarse image texture and isolated objects. Based on image texture and gross morphology, the mosaics have been classified into two categories of sediment (fine and coarse sand), seagrass beds, and four categories of reef (low relief reefs, patches, pinnacles, high relief continuous reef). The interpretation was ground truthed by independent diver observations and sediment grain size analysis. The pattern of reef development in these two leeward reefs is complex, with no shore-parallel zoning. The total spatial coverage of the reef classes is only 15% of the total surveyed area. Within these mapped reef areas, actual hard coral cover is estimated to be $<$ 50%, so the actual hard-ground coverage is relatively small. Within the various reef types, no simple relationship between morphological parameters such as spatial area, perimeter water depth and height was found. We speculate that the dominant control on reef development is antecedent granite or karst limestone topography.
H51C-1138 0800h
Quantifying Hillside Erosion Rates in California Over Varying Timescales
Quantifying erosion rates is important for predictive landscape modeling and helps to evaluate a landscape's response to forcing from climate, tectonic, and human impacts. Determining representative erosion rates is difficult, however, as rates vary spatially and temporally. Large episodic events like landslides and large floods may dominate over long timescales but often occur too infrequently to be recognized by monitoring programs over human time-scales. We present a quantification of erosion rates from a small (33 ha), well-studied basin in northern California using three independent methods operating on different timescales. Analysis of pond sediment volume from a trap at the base of the catchment yielded an average basin wide erosion rate equivalent to a rate of bedrock lowering of 87 +/- 20 m/Ma, over a 50 year period. Cosmogenic nuclide data for this site from Heimsath et al. (1997, 1999) indicates an average basin erosion rate of 102 +/- 25 m/Ma, for a timescale of 10,000 years. The use of the fallout radionuclides 137Cs and 210Pb to trace soil movement on the hillside yields a rough correlation of erosion rates with slope and curvature, supporting previously assumed transport laws. Rates range from 64 +/- 24 m/Ma on convex ridges to 264 m/Ma in hollows. The range of erosion rates determined using fallout nuclides agrees with rates from pond sediment volume and validates the use of our 137Cs and 210Pb to offer a calibration relationship for this site, enabling its application at other sites with similar characteristics. The good agreement between these three independent methods representing erosion rates over 40 to 10,000 year timescales suggests that the rates and processes observed today are dominant over long timescales as well. Such agreement validates assumptions made by past researchers at this site regarding temporal uniformity of process.
H51C-1139 0800h
137Cs Results and Interpretation of Cesium Soil Data on the Upper Fortymile Wash Alluvial Fan, Amargosa Valley, Nevada.
Studies using 137Cs were used to produce soil Cs profiles and to use them to determine erosion rates on interchannel divides of the Fortymile Wash alluvial fan over the last 50 years. Sample locations whose 137Cs profiles most resemble the reference-sample (stable surface) profiles are located on interchannel divide areas between distributary channels. These profiles are similar to the reference profiles that have low 137Cs values (in the range of 0.02 to 0.08 pCi/g) in the 3 to 6 cm layers. However, the surface layers (1-3 cm depth) typically have values much less than the reference samples from equivalent depths (range from 0.251 to 0.421 pCi/g). The data indicate that many of these interchannel divide areas have had part of the upper layer removed. Interchannel divide areas have the least likelihood of having been submerged during floods over the last fifty years. Thus, the loss of material from these otherwise stable surfaces appears to be due to eolian processes. Erosion of an interchannel divide area with little evidence of recent water movement is most easily explained by eolian removal. Evidence for wind erosion as the predominant process on the interchannel divide areas includes the lack of new or developing stream channels and the presence of modern coppice dunes near channels on interchannel divides. The presence of nearby Big Dune and other eolian deposits provides strong support for eolian erosion and transport. The amount of material removed from the interchannel divide areas was estimated by comparing the 137Cs value of the upper 3 cm layer to that of the reference value and calculating the thickness of the layer that would have to be removed to obtain the lower value. Applying this method across the interchannel divide sample locations indicates 1 to 2 cm of material has been removed from the interchannel divide surfaces in the last 50 years. This results in erosion rates that range from 0.02 to 0.04 cm/yr. These rates are similar to erosion rates of: (a) 0.019 cm/yr predicted to occur on farmland in Amargosa Valley (obtained from BSC 2004 [DIRS 169459], Section 6.4.2 by converting 0.19 kg/m2-yr using ash bulk density of 1 g/cm3; (b) 0.02 cm/yr estimated by the U.S. Department of Agriculture to have occurred on non-cultivated cropland and pastureland in Nevada (obtained from USDA 2000 [160548], Table 11 and calculated using 1 ton/acre-yr x 907 kg/ton x 2.47 x 10-4 acre/m x 0.001 m3/kg [bulk density] x 100 cm/m = 0.02 cm/yr). Overbank deposits on the interchannel divide areas indicative of periodic flooding are uncommon and restricted to narrow strips along the channel banks. The overbank and channel deposit samples have similar 137Cs signatures (the 3 to 6 cm layers and the 6 to 9 cm layers have nearly the same values in the 0.100-0.200 pCi/g range), indicating that the material from each environ was mixed during transport and deposited as a homogeneous sediment. The absence of many overbank deposits along the channel margins today indicates that flows sufficient to form extensive overbank flooding down Fortymile Wash and its distributary channels have not occurred in more than 50 years. Therefore, the channels currently transport most of their sediment load across the fan until it reaches the toe of the fan, where deposition occurs on the broad flats to the south or into the channel of the Amargosa River.
H51C-1140 0800h
Vegetation modulated landscape evolution: Effects of vegetation on landscape processes, drainage density and topography
Topography acts as a template for numerous landscape processes that includes hydrologic, ecologic and biologic phenomena. These processes not only interact with each other but also contribute to shaping the landscape as they influence geomorphic processes. We have investigated the effects of vegetation on known geomorphic relations, thresholds for channel initiation and landform evolution, using both analytical and numerical approaches. Vegetation is assumed to form a uniform ground cover. Runoff erosion is modeled based on power function of excess shear stress, in which shear stress efficiency is inversely proportional to vegetation cover. Plant effect on slope stability is represented by additional cohesion provided by plant roots. Vegetation cover is assumed to reduce sediment transport rates due to physical creep processes (rainsplash, dry ravel, and expansion and contraction of sediments) according to a negative exponential relationship. Vegetation grows as a function of both available cover and unoccupied space by plants, and is killed by geomorphic disturbances (runoff erosion and landsliding), and wildfires. Analytical results suggest that, in an equilibrium basin with a fixed vegetation cover, plants may cause a transition in the dominant erosion process at the channel head. A runoff erosion dominated landscape, under none or loose vegetation cover, may become landslide dominated under a denser vegetation cover. The sign of the predicted relationship between drainage density and vegetation cover depends on the relative influence of vegetation on different erosion phenomena. With model parameter values representative of the Oregon Coast Range (OCR), numerical experiments conducted using the CHILD model. Numerical experiments reveal the importance of vegetation disturbances on the landscape structure. Simulated landscapes resemble real-world catchments in the OCR when vegetation disturbances are considered.
H51C-1141 0800h
Challenges in Parameterizing a Landscape Evolution Model to Predict 1000 Years of Erosion on a Mesa-top Waste Repository
Low-level radioactive waste from operations at Los Alamos National Laboratory is currently being disposed in pits excavated into mesa-top repositories. One requirement for operation of the repositories is that no release of radioactive material will occur for up to 1000 years following closure of the repository. LANL is required to demonstrate that the repository can be successfully closed; including demonstrating that the waste pits will not be excavated by long term surface erosion processes. Here we describe the parameterization and application of the Siberia landscape evolution model in the optimization of the closure cap design for the Material Disposal Site G at LANL. The closure design is particularly challenging since some of the older waste pits are located near the edges of a slender finger mesa with complex topography. LANL scientists worked with the cap design engineer at URS in an iterative process to develop a stable design. In addition to design challenges, we faced significant challenges parameterizing the Siberia model for our site. Although we had multiple rainfall, runoff and sediment data sets at a range of scales (plot, first order basin and watershed), these data were not complete enough to develop the relationships required to parameterize Siberia. The data did however provide a strong realty-check on rainfall-runoff and runoff-sediment yield relationships derived from the IRS9 and Hillslope Erosion (HEM) models respectively, which were ultimately used to develop parameter values for the fluvial erosion in Siberia. Diffusion was constrained by eye-balling a match between Siberia generated topography and the observed topography (field and ALSM data) at the heads of first order channels. Siberia runs that generated colluvial hollows (non-existent at this site) indicated that the diffusion coefficient was too high. LA-UR-04-6328.
H51C-1142 0800h
Constraining Uncertainty in the Application of a Landscape Evolution Model to Predict 1000 Years of Erosion at a Mesa-top Waste Repository
A landscape evolution model (SIBERIA) is being applied to assess long term erosion of cap designs proposed for eventual closure of a waste disposal area atop a mesa, under the design requirement that waste remain buried beneath a prescribed minimum thickness of cap material at 1000 years. Here we describe a first-cut at constraining uncertainty in the landscape evolution model, so that predicted erosion rates are realistic within the bounds of the likely physical conditions at the site. A profile-based hillslope erosion, sediment transport, and erosion model (HEM) which has been subject to extensive validation on rangeland conditions is used to simulate sediment yield on hillslope profiles under a range of conditions that promote high, moderate, and low erosion rates. HEM uses a small set of parameters, including canopy- and ground-coverage, slope, and soil properties along the profiles, data for which are easily obtained in the field. Such a field dataset has been developed for profiles on disturbed and undisturbed hillslopes on the mesa and on an existing closure cap. The profiles included canopy- and ground-cover conditions which encompass the range of conditions expected to be present during the evolution of the ecosystem on the closure cap. A series of 64 HEM simulations were run with parameters obtained from the field dataset and properties of two soil types chosen to represent cap material properties on profile geometries constructed using a range of slopes. SIBERIA parameters were chosen so that resulting sediment yields on the profiles were bounded by the extremes of the HEM field based simulations. This represents a first step in assessing uncertainty in the application of SIBERIA as a predecessor to a more formal uncertainty analysis. LA-UR-6329
H51C-1143 0800h
Drainage Basin Erosion Rates Along the Death Valley Fault Zone, California From In-Situ Cosmogenic $^{14}$C in Alluvial Sediment: Preliminary Results
Catchment-wide erosion rates and drainage basin evolution are poorly understood processes in arid and hyper-arid climates. We collected samples from two small catchments in Death Valley, CA to determine basin-wide erosion rates using in-situ cosmogenic $^{14}$C to help understand drainage development in a hyper-arid setting. Death Valley is a tectonically active pull-apart system with a normal fault-bounded basin in the center and strike-slip faults to the north and south. On average, the area receives $<$ 5 cm of precipitation per year. Two basins were sampled from different parts of the fault zone; one along the normal fault-bounded segment in the Black Mountains and one from the southern strike-slip segment in the Owlshead Mountains. Topography along the dip-slip segment of the fault zone in the Black Mountains is youthful with steep, high relief catchments, triangular facets, and young fault scarps at the base of the range. Drainage basins in the Black Mountains have high ratios of basin excavated volume to planimetric area and high first-order stream gradients. Additionally, basins formed along the normal fault-bounded segment typically have large, 10-20 m high knickpoints and highly convex channel profiles. In contrast, the Owlshead Mountains are relatively low relief, and appear to be in a state of topographic decay. Samples were collected above and below a $\sim$20 m high knickpoint in the Badwater catchment, above and below a smaller (5 m high) knickzone in the Owlshead Mountains, and at the mouths of both catchments. In-situ $^{14}$C can be used to determine erosion rates in rapidly denuding areas that are integrated over a period of several thousand years. We are particularly interested in post-last glacial maximum erosion rates, which could be a function of climate change and the drying up of glacial Lake Manly, relative base-level fall resulting from recent movement on the fault zone, or a combination of both. Previous studies have shown that climate change since the last glacial maximum appears to have had little or no effect on erosion rates in eastern California. Because of this, drainage basin erosion rates may express a link between ground-rupturing earthquake activity on the Death Valley fault zone and denudation. The catchment-wide erosion rates may also help characterize fault zone segmentation, in conjunction with paleoseismologic, slip rate, and topographic metrics data.
H51C-1144 0800h
Implications of Groundwater Dynamics on Long-Term Changes of River Basin Topography and Hydrologic Response, With Application to the WE-38 Basin, Pennsylvania
Numerous studies have examined the impacts of geomorphology on the hydrologic processes of river basins, but much less attention has been given to the opposite problem: the effects of hydrologic processes on the evolution of river basin topography. Fluvial erosion processes are driven by surface runoff and streamflow, which depend on precipitation rates, soil moisture dynamics, and groundwater flow. All of these processes help determine spatial and temporal patterns of runoff and streamflow in a basin. Horton runoff occurs where the infiltration capacity is exceeded by the rainfall intensity and might produce relatively uniform incision across a basin. Dunne runoff and groundwater discharge typically occur in areas adjacent to river channels, thus eroding river networks and their neighboring locations. Groundwater is expected to be especially important to patterns of erosion when the infiltration capacity is large enough to absorb significant precipitation. On the Colorado Plateau, for example, groundwater leaves significant geomorphic signatures such as amphitheater-shaped channel heads and near constant valley widths from source to outlet. In this analysis, we investigate the role that groundwater movement plays in long-term landscape evolution using a landscape evolution model that has been modified to include a more detailed representation of basin hydrology. In the model, precipitation is generated by a stochastic process that includes realistic inter-storm variation, and the precipitation is partitioned between surface runoff and groundwater recharge using specified infiltration and recharge rates. Groundwater flow is simulated by a two-dimensional Dupuit equation for a homogeneous, isotropic, unconfined aquifer with an irregular underlying impervious layer. The model is applied to the WE-38 basin, an experimental watershed in Pennsylvania. This site was selected as a study area because substantial hydrologic and geomorphic information is available including rainfall data, streamflow data, groundwater table elevations, and estimated parameters for the geomorphic processes. First, the hydrologic model is calibrated to match observed streamflow data, then the combined hydrologic/geomorphic model is used to investigate several hypothetical scenarios. The scenarios investigate the role of groundwater as the infiltration capacity, hydraulic conductivity, and impervious layer are modified. The resulting topographies are analyzed and their hydrologic behavior is characterized. The results indicate that groundwater plays an important role in shaping fluvial landscapes and thus affects the long-term evolution of hydrologic response, especially in basins with large infiltration capacities and thick aquifers.
H51C-1145 0800h
Geomorphology of Beach Ridges and Holocene Terraces on Kamchatka: A Complex Interplay of Tectonics, Volcanism and Coastal Processes
Despite decades of analysis, the origin of beach ridges remains debatable, and coastal morphotectonics an open field. On the eastern coast of Kamchatka, along about 700 km of coastline (northwest Pacific and southwest Bering Sea), we have measured accumulative Holocene coastal profiles spanning a range of conditions including variable tectonics (co-seismic and longer term), wave climate, sea ice, and sediment supply (including fluvial, debris-flow, eolian, longshore transport). We can constrain the age of many of these coastal features (beach ridges, terraces) using well-mapped and dated Holocene marker tephra-typically present at time intervals of 500-1000 yrs (or less) over several thousand years. Here are some generalizations for Kamchatka's east coast: 1) Beach ridge complexes are typical of coastlines that are relatively stable tectonically (little or no uplift, mild subsidence); coastlines undergoing uplift at rates on the order of 10-100 cm/kyr exhibit terraces, rather than beach ridges. Subsiding coastlines are erosional. 2) Some beach ridges are clearly correlative with co-seismic subsidence (local sea-level rise) and erosion, building up and then stranding a beach ridge during the interseismic cycle. Other beach ridge complexes are well-developed in areas where co-seismic deformation is unlikely. 3) Beach ridge recurrence intervals range from decades (Uka site in the Bering Sea) to hundreds of years (e.g. south Kamchatka). Especially in narrower embayments, beach ridge development in maximized toward the center of the bay. 4) Along the Kamchatka coast, beach ridge height is highly variable, and in some cases tied to eolian processes, or age duration. Beach ridge height on any one profile is also variable. If the shoreline is subsiding, not prograding, or eroding, the first beach ridge may be two times higher than the ridges landward.
H51C-1146 0800h
Interactions between Vegetation and Land Surface Evolution in Arid and Semiarid Systems
Ecological, hydrological and geomorphological processes are tightly coupled and the understanding of their complex interactions represents a research challenge that is at the heart of the emerging fields of ecohydrology and ecogeomorphology. The coupling between ecology and hydrology is particularly strong in drylands, in which water limited conditions are the main constraint for vegetation growth and survival. Arid and semiarid ecosystems comprise about 30$%$ of the Earth's surface. Various forms of environmental perturbations like climate change and anthropogenic activities can lead to desertification or degradation of these ecosystems. The vegetation of water-limited ecosystems is commonly patterned, that is, arranged in a two phase mosaic composed of patches with high biomass cover interspersed within a low-cover or bare soil component. These patterns play an important role in controlling erosion. Human impacts or climate change may alter these systems, disrupting vegetation and triggering erosion. The resulting geomorphic changes are likely to have feedbacks including runaway desertification. Models that couple erosion and vegetation evolution can be used as a tool to understand the dynamics of arid and semi arid systems and the impact of climate change and human disturbance. However, the few existing models that explore the interactions between vegetation and sediment movement do not account for the feedbacks between ecohydrologic and geomorphic processes as is the objective of our research. A new modeling framework that couples the SIBERIA landform evolution model with a dynamic vegetation model for water limited ecosystems will be presented. The model explicitly accounts for the dynamics of runon-runoff areas that controls the evolution of the spatial distribution of vegetation in water limited ecosystems. The model reproduces the dynamics of banded vegetation patterns (tiger bush) characteristic of areas with mild slopes as well as more complex two-phase patterns that are characteristic of steeper areas. Complex spatially distributed feedbacks into slope evolution are also apparent. Preliminary results on the long term topographic evolution of hillslopes with these different dynamic vegetation patterns as well as the geomorphic consequences of disrupting or eliminating the vegetative cover will be presented.
H51C-1147 0800h
Soil Production, Landscape Evolution and Vegetation Dynamics in the Blue Mountains, Australia.
Soil production is thought to relate to overlying soil depth by an inverse exponential function, as empirically derived using terrestrial in situ cosmogenic nuclides (TCN) at several study areas. This contrasts with a long held assumption that soil production is maximised under a thin soil cover, dm (Gilbert, 1877). Many sites in the Blue Mountains, Australia, display prima facie morphologic evidence for a `humped' soil production function. A sharp soil depth change occurs between proximal and distal spur extremities that accord with a change from forest to treeless heath, and shallow discontinuous soils are found on spur noses. Either of these features may indicate unstable conditions at depths less than dm implicit in a humped soil production function. We attempt to constrain the soil production function at our site using the TCN Be-10 from sandstone bedrock and saprolite, and morphometric analysis at Marrangaroo Creek. Although the soil depth change from forest to heath may be related to regional curvature not soil production, the peak in soil production under shallow mantles may explain alternating bands of soil and outcrop on spur noses. Soil production rates are mildly influenced by overlying soil thickness, suggesting that although thin layers of iron cemented sandstone only comprise a small percentage of the catchment bedrock, its resistance to weathering sets the pace of surface lowering. Furthermore, we present evidence that Marrangaroo Creek is a result of post-Miocene incision, similar to adjacent catchments in the region.
H51C-1148 0800h
Modeling Ecologic and Geomorphic Change in Semi-Arid New Mexico
A record of grazing, fire suppression and drought in the semi-arid American south-west since the 1800s has left a mark on the landscape. In the Pajarito Plateau, New Mexico, pinyon-juniper woodlands have expanded at the expense of ponderosa pine forest, giving rise to increased erosion rates - a concern to the sustainability of the soil resource and of vegetation. Soil loss also risks human exposure to contaminants from Los Alamos National Laboratory. The dynamics of this water-controlled environment arise from the complex interplay of climate, hydrology, ecology and geomorphology, an understanding of which is fundamental to solution of the environmental challenges. In an effort to predict the course of ecological and geomorphic change in the region, and to provide a tool for land-use planners in averting possible human health and environmental hazards, a numerical landscape evolution model is developed. It assimilates research encompassing plant physiology, ecology, hydrology, and geomorphology, highlights weaknesses in process understanding, and provides a prediction of the future environmental state on the time-scale of 100s of years.
H51C-1149 0800h
Effects of a Wildfire and Salvage Logging on Hillslope Erosion: Star Fire, Placer County, California
Post-fire erosion rates have been measured in many areas, but there are few published data on how salvage logging affects post-fire erosion. The primary objective of this study was to compare sediment production rates from sites burned at high severity and subjected to helicopter, cable, or tractor logging, respectively. Sediment production was measured with sediment fences on 32 logged sites and five unburned, unlogged sites over two wet seasons in the central Sierra Nevada of California. The independent variables measured on each site included slope, aspect, contributing area, percent bare soil, percent rocky outcrop, percent ground disturbance, soil texture, and soil water repellency. The first wet season had near-normal precipitation and a calculated erosivity of 556 MJ mm ha$^{-1}$ h$^{-1}$. Mean sediment production rates were 2.5 Mg ha$^{-1}$ from the burned sites and zero from the unburned sites. The second wet season had only 13 MJ mm ha$^{-1}$ h$^{-1}$ of erosivity because most of the precipitation fell as snow, and the mean sediment production rate from the burned sites declined to 0.11 Mg ha$^{-1}$. Mean sediment production rates did not significantly differ by logging type due to the high variability between sites in the first wet season and the very low sediment production rates in the second wet season. Slope, percent ground disturbance, and percent bare soil were significantly correlated with increasing sediment production per unit area. Percent ground disturbance was significantly higher for the cable- and tractor-logged sites than for the sites logged by helicopter (p$<$0.001), but there were no significant differences in percent bare soil by logging type. Sites logged by helicopter tended to have lower sediment production rates than cable-logged sites (p=0.086). The results indicate that logging methods that increase ground disturbance and bare soil will generate more sediment, but statistically significant differences in sediment production may be difficult to detect given the natural variability between sites and the variability in harvest practices.
H51C-1150 0800h
Predicting Tectonics From Topography: Case Study Using SRTM Data From the Namche Barwa Region, Eastern Himalayan Syntaxis, Tibet
The eastern Himalayan syntaxis, Tibet, is a tectonically active region located at the `corner' of two colliding plates. As such, it is an ideal location in which to use NASA SRTM-derived topographic metrics to elucidate surface-tectonic interactions, especially around the rapidly exhuming Namche Barwa massif. In order to minimize the effects of differing lithology, the extraction of topographic indices was limited to basins draining the medium- to high-grade metasediments and gneisses on the west and north sides of Namche Barwa. Landsat TM data calibrated with field observations are used to identify the difference between bedrock and alluvial channels, as well as which portions of the landscape are controlled by glaciers and the difference between bedrock and alluvial channels. The Nyingqui River is a $\sim$SE-flowing, alluvial channel at its confluence with the NE-flowing Tsangpo. Elongate basins draining into the NE side of the Nyingqui trend NE-SW and average $\sim$200 km$^{2}$. Rounded basins that drain into the Tsangpo immediately downstream of the confluence trend N-S to NW-SE and are smaller, $\sim$60 km$^{2}$. Two basins that lie over the drainage divide to the north trend N-S and E-W and average $\sim$1500 km$^{2}$. The rivers draining into the Nyingqui exhibit convex-up long profiles, and their associated hypsometric curves indicate that a large percentage of this area is concentrated at higher elevations, well above the Nyingqui's base level. Asymmetry in these NE-SW trending basins and in the first basin downstream of the confluence indicate tilting to the NW. These trends change dramatically downstream of the Tsangpo confluence, where the long profiles range from graded to straight, hypsometry indicates more even distributions of elevation within the basins and asymmetry favors tilt to the E or NE, depending on the orientation of the basin. Asymmetry of the larger basins over the divide suggest tilt in opposite directions, to the SE and the NW, respectively. Regional topographic residual and slope maps indicate that the basins draining into the Nyingqui and those on the other side of the divide have low local relief and slopes relative to the smaller basins that feed the Tsangpo. The systematic change of these metrics from basin to basin suggests recent tectonic influence, specifically regional uplift and tilting of this region to the NW, consistent with a network of well-constrained, low-temperature thermochronologic data (Malloy et al., 2003) by the larger NSF Geodynamics of Indentor Corners project.
H51C-1151 0800h
Applications of Synthetic Aperture Radar Images to Map Geologic Structures and Geomorphology: Evaluating the Interaction of Tectonic and Surface Processes in the Saint Elias Mountains, Alaska
The Saint Elias Mountains of southern Alaska and Canada are forming by collision of an allochthonous terrane at the northeastern end of the Aleutian subduction zone. The region represents one of the most dynamic glacial and tectonic systems on earth. Remote sensing is an important tool for interpreting structural and geomorphic features and processes in this remote and glaciated region of steep terrain and high relief where vegetation ranges from dense coastal conifer forest to alpine tundra. We use terrain corrected SAR amplitude images (brightness and radar backscatter) from the C-band ERS 1,2 and RADARSAT-1, and L-band JERS-1 imaging radars to evaluate the use of satellite borne radar for mapping structural and geomorphic features. Radar images are terrain corrected and draped upon 30 m posted digital elevation data obtained from the Shuttle Radar Topographic Mission flown in 2000. We have also written visualization software to interactively extract orientation data of structural features from 3D images including strike and dip of sedimentary rocks, faults, axial surfaces of folds, and contacts defined by angular unconformities. Geomorphic features of interest include fault scarps, braided river deposits, constructional and tectonically uplifted beach berms, wave cut terraces, rock and land slides, and glacial features including crevasse systems, moraines, and thermokarst topography. Orientation of crevasse systems on glaciers is automatically extracted using an algorithm based upon the directional anisotrpy of the 2D Fourier transform of radar images. The rate of transport of rock slide debris by glaciers is measured with a feature tracking algorithm applied to time-sequence imagery. L-band radar is superior to C-band for mapping geomorphic features of low topographic relief in densely vegetated lowlands. These include beach berms, relic river drainage patterns, and vegetated glacial topography. Radar response of structural features in bedrock is mostly controlled by hill slope and aspect, which produces similar response in radar brightness on both L- and C-band data in areas of modest to low vegetation density. We are continuing to evaluate the effects of radar polarization (HH vs VV), variability in beam incidence angle, and ascending and descending orbits on feature identification.
H51C-1152 0800h
Morphometric Classification and Multivariate Analysis of Drainage Networks in Arid Rift Basins Using Remote Sensing Data, Baja California, Mexico
Drainage networks in continental rift basins are strongly linked to bedrock lithology, climate, and fault systems, although quantitative analyses that demonstrate these relationships are generally lacking. We used SRTM data and enhanced TM+ images from Baja California, Mexico to investigate the morphometric characteristics of drainage basins along the eastern side of the peninsula, which is typified by numerous rift basins of late Miocene age. These basins are now in their early post-rift stages of evolution and are located in a similar climatic zone along the entire length of Baja, although bedrock lithology and hypsometry for each basin is drastically different. As a first step, SRTM data, enhanced TM+ images, ASTER images, and geostatistical techniques (cokriging methods and Alternating Conditional Expectations algorithms) were used to construct digital terrain models (DTM) with 15-meter spatial resolution for nearly the entire eastern rift province of Baja. Watersheds were extracted from the DTM, which then allowed the uniqueness of each drainage basin to be characterized quantitatively using various basin attributes and morphometric indices. Each basin's geomorphic development and the topographic forcing factors responsible were then examined and classified using multivariate statistics. Hypsometry for each basin was generated and found to be critical for evaluating the "tectonic maturity" of each basin. More specifically, the hypsometric integral is sensitive to the recent tectonic history of each basin. Multivariate analyses (Principal Component and Cluster Analyses) were also used to classify Baja rift basins into similar geomorphometric groups.
H51C-1153 0800h
Downslope Asymmetry in Rainsplash Transport of Sand Revealed With High-Speed Imaging
An understanding of rainsplash detachment and transport is required to formulate theories for a host of hillslope processes, including downslope rainsplash transport, rill development and destruction, and sheetwash transport resulting initially from rainsplash. Net downslope transport of soil particles by rainsplash results from a bias in the number and/or length of particle trajectories. The relative contributions of these attributes of particle motion bear on formulations relating rainsplash transport rates to surface slope angle, as the directional asymmetry of trajectories depends on details of motion during drop impact, including drop incidence angle, whereas the bias in trajectory lengths mostly obtains for geometrical reasons. To clarify these attributes of particle motions initiated by rain drop impacts, we are conducting experiments that allow us to directly visualize them. Specifically, in a first set of experiments we have obtained high-speed video imagery of water-drop impacts and associated motions of medium sand particles on varying slopes under dry and moist (drained) conditions. Drops with diameters of 3 mm were released from a height of 5 m onto a sand target. The target consisted of a circular hole (diameter of 2.5 cm, depth of 1.9 cm) drilled into a wooden block. The quartz sand was angular with a nominal diameter 0.35 mm. In each run the sand surface was initially smooth and flush with the surrounding target surface. Moist conditions were obtained by filling the target hole with sand, wetting it, then allowing the sand to fully drain. Target slopes were 0, 10, 20 and 30 degrees. Images of the drop impacts and particle motions were obtained with a high-speed digital video camera running at 500 frames per second. The videos reveal a clear radial symmetry in particle trajectories at 0 degrees, and only a slight asymmetry at 10 degrees. Noticeable asymmetry occurs at 20 degrees and, at 30 degrees, virtually no particles move upslope. During drop impact the mass and momentum of the drop spreads laterally, causing a surface layer of particles (several particle diameters thick) to move radially outward as much as several drop diameters. This motion is strongly asymmetrical at larger slopes and, with dry conditions, may be as important as ballistic motions in producing net transport. Dry conditions involve more ballistic motions than do moist conditions. The effects of surface tension associated with initially moist conditions are to reduce the detachment of surface grains, reduce plowing and, for particles that do detach, increase the likelihood that particles cling together during their ballistic motions. These results suggest the possibility of decreasing rainsplash transport with initial surface wetting following the onset of a storm, followed by increased transport as surface-tension effects are reduced with further wetting toward saturation, before the onset of rainsplash-enhanced particle mobilization with sheetflow. The results also suggest that the drop incidence angle, set by both surface slope and drop trajectory, contributes significantly to asymmetry in particle trajectories. Clarification of these points, however, will require experiments that cover a wider rage of parametric quantities (e.g. drop size, particle size, moisture content, surface roughness) than those considered in these initial runs.
H51C-1154 0800h
Predicting Cumulative Watershed Effects using Spatially Explicit Models
Cumulative watershed effects /(CWEs/) result from the combined effects of land disturbances distributed over both space and time. They are of concern because changes in flow and sediment yields can adversely affect aquatic habitat, channel morphology, water yields, and water quality. The assessment procedures currently used by agencies such as the U.S. Forest Service generally rely on a lumped approach to quantify disturbance, despite the widespread recognition that site conditions and location do matter! The overall goal of our work is to develop spatially-explicit models to quantify changes in flow and sediment yields. Key objectives include: use of readily available GIS data; ease of use for resource managers with minimal GIS experience; modularity so that models can be added or updated; and allowing users to select the models and values for key parameters. The DeltaQ model calculates changes in peak, median, and low flows due to forest management activities and fires. Inputs include GIS data with disturbance polygons, an initial change in flow rate, and the time to recovery. Data from paired watershed studies are provided to help guide the user. The initial version of FORest Erosion Simulation Tools /(FOREST/) calculates sediment production from forest harvest, fires, and unpaved roads. Additional modules are being developed to deliver this sediment to the stream channel and route it to downstream locations. In accordance with our objectives, the user can predict sediment production rates using different empirical equations, assign an initial sediment production rate and a specified linear recovery period, or develop a look-up table based on local knowledge, published values, or data from other models such as WEPP. The required GIS layers vary according to the model/(s/) selected, but generally include past disturbances /(e.g., fires and timber harvest/), roads, and elevation. Outputs include GIS layers and text files that can be subjected to additional analyses. Both DeltaQ and FOREST will be demonstrated and can be downloaded at: http://www.cnr.colostate.edu/frws/people/faculty/macdonald/macdonald/model.htm.
H51C-1155 0800h
Topographic Signatures of Geomorphic Processes at Desert Piedmonts
Geomorphic processes operating on desert piedmonts leave their mark in the topography of the piedmonts, measured at a range of scales. For example, at the km to 10's of m scale, slope, curvature, and dissection give clues to the relative importance of erosional, aggradational, and transport processes. When a desert piedmont is dominated by erosional and transport processes, a pediment is formed. Conversely, when aggradational processes dominate, alluvial fans are formed. The relative rates of uplift vs. aggradation determine the slope of the fan and whether the fan is dissected. Changes in uplift rate or climatic conditions can lead to isolation of the currently forming fan surface through entrenchment and construction of another fan either further from the mountain front (decreased uplift or increased runoff) or closer to the mountain front (increased uplift or decreased runoff). Fitting 3-dimensional surfaces to alluvial fan units allows parameters for the apex position, slope, and radial curvature to be compared with unit age. The topographic signature of individual fan units will help in regional correlation of fan surfaces and may assist in separating the effects of climate and uplift. At scales of m to cm, processes such as salt weathering, aeolian deposition, and desert pavement formation dominate. The lithologies that make up the piedmont determine their susceptibility to processes of disintegration due to salts (e.g. coarse-grained rocks) or dissolution (e.g. limestones). The composition of the source areas also helps determine the relative importance of debris and mud flows. Resistant lithologies tend to form desert pavement surfaces with time, caused by infiltration of aeolian dust below a monolayer of resistant pebbles to cobbles. This leads to local relief of a few cm, while surfaces composed of easily weathered lithologies have much lower relief at that scale. Work performed under contract to NASA.
H51C-1156 0800h
Erosion Rates, Landscape Morphology, and Hillslope Processes in the Upper Beni River Region, Bolivian Andes
Over the long term, rates of mountain erosion are controlled by rates of channel incision into bedrock. Channel incision rates are a function of discharge and local channel gradient, which reflects lithology and rock uplift rate. Other things being equal, patterns of channel gradient are indicators of relative channel incision rates. One useful metric of relative gradient is channel steepness index, k$_{s}$, the coefficient modifying a power law relationship between local channel gradient and contributing drainage area. In the Upper Beni River region of the Bolivian Andes, short-term, basin-averaged erosion rate correlates with channel steepness index. Erosion rates derived from analysis of in situ-produced $^{10}$Be in alluvium range from 0.05 mm/yr to 1.35 mm/yr, and average over 10$^{2}$-10$^{4}$ years. Channel steepness index values, for a reference concavity of 0.45, are 100-200 in the headwaters of trans-range drainages, increase to 400-600, and then decrease to 40-100 downstream. Departures from this pattern occur in two channel networks with exceptionally large drainage areas whose headwaters have extended into the internally drained Altiplano on the Andean plateau. Patterns of channel steepness index, and by implication incision rate, appear primarily to reflect tectonic patterns and transient adjustment to those patterns by channel networks. Lithology exerts a weaker control on erosion rate but influences spatial patterns of hillslope geomorphic process. Neither channel steepness index nor basin-averaged erosion rate shows strong correlation with mean basin hillslope gradient or mean basin local relief because many hillslopes in the Upper Beni River region are at threshold values of slope and local relief. In addition, some hillslopes appear to be in a transient state of adjustment to relatively rapid rates of channel incision.
H51C-1157 0800h
A Numerical Simulation of the Effects of Mass-Wasting on Cosmogenically Determined Erosion Rates
The successful quantification of long-term erosion rates underpins out understanding of landscape formation, the topographic evolution of mountain ranges, and the mass balance within active orogens. The measurement of in situ-produced cosmogenic radionuclides (CRNs) in fluvial and alluvial sediments is perhaps the method with the greatest ability to provide long-term erosion rates over a wide variety of landscapes and erosional processes. Recent studies have provided encouragement that measurement of in-situ CRNs does, in fact, frequently yield consistent long-term average erosion rates. Cosmogenically derived erosion rates from small catchments within the Sierra Nevada are highly consistent (e.g. Riebe et al., 2000), while CRN samples collected along European rivers yield similar results at catchment scales of 10$^2$-10$^4$ km$^2$ (Schaller et al., 2001). On the contrary, samples from low-order catchments in the Nepalese Himalaya yield cosmogenically derived erosion rates that range from hundredths to tens of mm/yr (A. Heimsath, unpub. data), while results reported from the San Bernardino Mountains in southern California yield erosion rates that vary from $<$0.01 to $>$2.7 mm/yr (Binnie et al., 2003). We propose that stochastic mass-wasting processes are responsible for the wide range of cosmogenically derived erosion rates observed in active tectonic settings, and have developed a numerical simulation of cosmogenic nuclide production and distribution in landslide-dominated catchments to address the effect of these mass-wasting processes on cosmogenic erosion rates in active landscapes. This model simulates the production of cosmogenic nuclides on the landscape, and their removal from the landscape by erosional processes, such as grain-by-grain attrition and mass wasting. CRN concentrations in the `removed material' are analyzed at different catchment scales to yield simulated cosmogenic erosion rates. Results of such simulations indicate that the temporal stability of erosion rates determined from CRN concentrations in alluvial sediment decreases with increased ratios of mass-wasting to grussification rates within a given catchment area, and that increasingly large catchment areas must be sampled when mass wasting rtaes are high in order to accuately evaluate long-term erosion rates. In addition, results of this simulation suggest that sediment sampling for CRNs is the appropriate method for determining long-term erosion rates in regions dominated by mass-wasting processes, while bedrock surface sampling for CRNs is generally an ineffective means of deriving long-term erosion rates.
H51C-1158 0800h
The Effects of Hyperaridity on Soil Production and Transport on Hillslopes: Adapting Geomorphic Models in the Atacama Desert, Chile
On most soil-mantled hillslopes, slope-dependent sediment transport is primarily driven by biological processes and, at steady-state, the eroded soil is replaced by production from the underlying saprolite. We are investigating how soil production, weathering, and transport change as precipitation and biological activity decrease to near zero values in order to better understand soil and geomorphic processes in hyperarid deserts and on Mars. We are studying convex hillslopes in the Atacama Desert, Chile along a precipitation gradient (~55 mm to ~2 mm yr$^{-1}$). All sites have similar bedrock, elevation, and temperature, and are surrounded by fluvial sediments of Pliocene age (based on $^{10}$Be and $^{26}$Al exposure ages and Ar-Ar dating of volcanic ash). Our field observations suggest that as precipitation declines to near zero, biologically driven sediment transport ceases and atmospheric deposition may replace saprolite erosion as the dominant mode of soil production. Here we report our initial results for the rates of saprolite erosion and surface gravel transport on these hillslopes. We also present a hillslope soil mass balance model for the Atacama Desert, based on work by Heimsath, Dietrich, and others, that it explicitly includes atmospheric deposition and chemical processes as controls on hillslope soil thickness.
H51C-1159 0800h
Identifying Watershed Sediment Sources In The Chesapeake Bay
Attenuation of light by fine-grained suspended sediment is having an adverse affect on the living resources and habitat of the Chesapeake Bay and its watershed. Different approaches are being used to identify sediment sources at several scales for the Chesapeake Bay watershed. At the subbasin scale (1.0 to 70,200 km $^{2}$), U. S. Geological Survey suspended-sediment data from 1985 through 2001 for 35 stations with at least 3 years of record were used to determine subbasin sediment yields. In the Susquehanna River Basin results showed that four streams draining the Conestoga River Basin,(1,220 km $^{2}$) which is in the Piedmont, had the highest sediment yields (60.9 to 356 t/km $^{2}$/yr). Cosmogenic $^{10}$Be provides another method to measure erosion which can be compared to subbasin sediment yields. Two pathways of the cosmogenic radionuclide $^{10}$Be, atmospheric and {\it in situ}, were used to determine erosion rates in the Susquehanna River Basin (70,200 km $^{2}$). Atmospheric $^{10}$Be was used to generate erosion indices at 25 subbasins by taking a ratio of $^{10}$Be in fluvial sediment exported out of the subbasin against the net atmospheric delivery of $^{10}$Be (values $>$1 = erosion). Examining the relation of {\it in situ}$^{10}$Be concentrations compared to subbasin sediment yield provided an independent method to assess instrumental vs. background erosion rates. Subbasins in equilibrium show a linear relation of instrumental sediment yield to {\it in situ}$^{10}$Be concentrations. Subbasins that deviate from this relation show either export or storage of sediment. Subbasins of the Conestoga River Basin showed departure from this relation, indicating erosion. The Conestoga River Basin drains primarily agricultural land and this land use may be influencing erosion rates and sediment yields. Within Chesapeake Bay subbasins, sediment fingerprinting is being used to determine watershed sources of sediment. Sediment fingerprinting is a technique where potential sediment sources can be characterized using a number of diagnostic physical and chemical properties. Results from the Pocomoke River Basin near Willards, MD (USGS ID 01485000)(157 km $^{2}$) for 7 storms during 2001-02 showed that the channel corridor (bed and banks) was the major source of suspended sediment, contributing between 61-100%. Cropland contributed between 0 and 39%. Significant fingerprints used in this analysis included $^{137}$Cs, $^{13}$C, and $^{15}$N. Erosion mass balance using $^{137}$Cs also confirms that cropland erosion was low.
H51C-1160 0800h
Sixty Years of River Corridor Change Induced by the Construction, Operation, and Maintenance of Flood Control Infrastructure on Lower Deer Creek, California
Deer Creek drains 540 km2, joining the Sacramento River near Vina, about 160 km north of the city of Sacramento. The U.S. Army Corps of Engineers constructed a levee and partly straightened the lower five miles of Deer Creek in 1949. Repeated levee failures and the presence of the federally threatened spring-run Chinook salmon (Oncorhynchus tshawytscha) in Deer Creek have prompted investigations on habitat restoration coordinated with more effective flood protection. The Deer Creek Watershed Conservancy (1998) identified a significant reduction in channel complexity between 1938 (pre-levee) and 1997, but did not attempt to quantify this reduction. In this study, we examined high quality aerial photographs from 1938, 1952, 1966, 1979, 1985, and 1998, and systematically quantified (in ArcGIS) changes in river corridor complexity by digitizing a range of features in each set of photos. Total active channel length in the levee reach decreased from 14.4 km to 12.6 km between 1938 and 1998. In addition, we documented a significant increase in average active channel width and a decrease in shaded riverine aquatic habitat between 1938 and 1998. Most of these changes occurred during the levee project in 1949, and the simplified channel form persisted through 1998. We also identified a significant decrease in aquatic and riparian habitat resilience (i.e. resistance to habitat damage and destruction by large floods) between 1937 and 1998. These results provide a basis for prioritizing, locating, and developing designs for alternative flood management approaches that would contribute to the enhancement and restoration of aquatic and riparian habitat along lower Deer Creek.
H51C-1161 0800h
Geomorphic Assessment Approach to Evaluate Stream Channel Stability for Regions of Illinois, Case Study: Southern Illinois Region
An array of different geomorphic assessment approaches for evaluating stream-channel stability is being utilized throughout the country to meet the demands of resource managers interested in stream channel restoration and management to reduce erosion and improve stream habitat. Over the last century, most of the Illinois landscape has experienced intensive land use changes which have contributed to stream channel instability. Stream channels in Illinois have adjusted to these changes either by increasing lateral rates of migration, downstream translation of meanders, widening, or development of headward retreat of knickpoints, depending on the region of the state. Illinois can be divided into at least four regions based on prevailing physiographic features and style of channel adjustment. Also, channel response in most of these regions tend to be more subtle than the dramatic response characteristics of streams in the Coastal Plains, mountain environments, and the desert southwest for which other geomorphic approaches have been developed. The observed magnitude and type of channel response are related to topography of the bedrock surface and extent and morphology of several glacial periods, which carry local significance for stream management. Given that geomorphic assessments for stream restoration require non-trivial professional, time, and financial resources, the development of approaches for Illinois regional conditions are more beneficial. A geomorphic assessment approach is being developed by adapting methods from existing process-based approaches utilized around the United States. A case-study was performed in the Big Creek watershed of the Cache River Basin for the southern Illinois region. This region was selected first because it exhibited dramatic channel responses to disturbances and had an extensive hydrologic, sediment, and land management record. This adapted approach includes systematic data collection protocols for characterization leading to an evaluation of the fluvial system for the purpose of determining past watershed and channel conditions, current geomorphic character, and potential for future channel adjustments using modest time and financial resources. The performance of this approach in the southern Illinois region and future development for other Illinois regions will be presented.