H53C-1377
Does Rock Mass Strength Control the Rate of Alpine Cliff Erosion?
Collapse of cliff faces by rockfall is a primary mode of bedrock erosion in alpine environments and plays a controlling role in mass removal from these systems. In this work we investigate the influence of rock mass strength on the retreat rate of alpine rock slopes. To quantify rockwall competence we employed the Slope Mass Rating (SMR) geomechanical strength index, which combines numerous factors that affect the strength of a rock mass, such as intact rock strength, joint frequency, joint condition, and more. The magnitude of cliff retreat was calculated by estimating the volume of talus at the toe of each rockwall and projecting that material back onto the cliff face, while accounting for the loss of production area as talus buries the base of the wall. Selecting sites within basins swept clean by advancing LGM glaciers allowed us to estimate the time period over which talus accumulation occurred (i.e. the production time). Dividing the magnitude of normal cliff retreat by the production time, we calculated erosion rates for each site. Our study area included a portion of the Sierra Nevada from Yosemite National Park in the south to Lake Tahoe in the north. Rockwall recession rates determined for 40 alpine cliffs in this region varied from 0.02 to 1.22 mm/year, with an average value of 0.28 mm/year. We found good correlation between rockwall recession rate and SMR that is best characterized by an exponential decrease in erosion rate with increasing rock mass strength. Analysis of the individual components of the SMR reveals that joint orientation (with respect to the cliff face) is the most important parameter affecting the rockwall erosion rate. The complete SMR score, however, best synthesizes the lithologic variables that contribute to the strength and erodibility of these rock slopes. Our data reveal no strong independent correlation between the measured rockwall retreat rate and environmental attributes (such as site elevation, aspect, cliff slope length, and cliff slope angle), suggesting that rock mass strength is the dominant parameter controlling the rate of cliff erosion in our study area. http://www.ce.berkeley.edu/~moore/projects/talus.html
H53C-1378
Rock Matters: Lithologic Controls on Landscape Evolution
Landscapes are shaped through balances between external forcing (climate, tectonics, humans) on erosion and internal controls like material strength. It is widely held, and commonly assumed, that erosion rates vary directly with tectonic uplift rates. It is, however, still relatively unclear how the balance between tectonic forcing and lithologic properties impacts landscape evolution. We focus this study to explore whether landscapes of different rock types evolve differently under similar tectonic forcing. Our field area is the San Gabriel Mountains in southern California. The entire mountain range falls within a similar climatic zone, and it has both a tectonic and lithologic gradient. Previous work quantified a clear correlation between catchment steepness, erosion rates and inferred uplift rates across the field site. For work presented here we chose three lithologies (anorthosite, granite, and mixed metamorphics) across various erosional-uplift regimes, from low-relief soil mantled hill slopes to high- relief rocky hill slopes. At each site we measured parent material strength with a drop-cone penetrometer and/or a shear vane tester. Preliminary analyses suggest that material strength varies significantly with relative erosion rate. We also show that material strength does not vary significantly between rock types under similar erosional regimes. Because we observe quantitative differences in material strength for a given lithology under different erosion rates, we show support for the hypothesis that erosion rates influence material strength and not the reverse. Keeping in mind that the only other variable considered here is tectonic forcing, we suggest that relative erosion rates are controlled more by tectonic forcing than lithology, but that lithology may still account for the variability within an erosional regime.
H53C-1379
Tectonic and Climatic Controls on Landscape Development of Puerto Rico
The northeastern Caribbean island of Puerto Rico is an exhumed Cenozoic island arc situated between the inactive Muertos trench to the south and the highly oblique Puerto Rican Trench to the north that forms the left- lateral strike-slip plate margin with North America. The rectangular island's long axis of 175 km parallels the east trending strike of the trenches with a near constant width of between 50 and 60 km. Puerto Rico receives the NE trade winds and has a tropical monsoonal climate. Puerto Rico has a distinct midline asymmetry with north draining watershed about twice the length and five times as large as south draining watershed. This midline asymmetry is more pronounced along the islands eastern third than the central or western thirds. River outlet spacing, mountain front sinuosity, and comparative hypsometry display similar east to west variability consistent with greater denudation in the eastern parts of the island. The southwestern fifth of the island is underlain by serpentinized ocean crust that forms the large diapiric Monte del Estado uplift. Active diapirism is indicated by highly asymmetric watersheds of the surrounding rivers and tributaries. Stream length gradient index calculated from 1:20,000 scale map data and compared to fault locations show little correlation suggesting that active faults does not significantly control Puerto Rico's landscape. Quantified morphologic data from the eastern two-thirds of Puerto Rico are consistent with a landscape developed in response to the precipitation derived from NE trade winds while serpentinite diapirism dominates the western third of the island. Individual active faults of Puerto Rico do not control the landscape development.
H53C-1380
Numerical modelling of glacial landscape response to tectonic and climatic forcing
Glacial erosion represents a potentially crucial coupling between tectonic processes and climate change. However the role of glaciers in this context remains little explored. Previous digital topographic analyses have suggested that the response of glaciers to tectonic forcing is strongly dependent on glacier size; large glaciers behave as an effective glacial "buzzsaw", while small glaciers behave much like rivers and steepen in response to rock uplift. A second striking feature is the development of tall cirque headwalls in association with rapid rock uplift. Numerical modelling of glacial longitudinal profile evolution allows us to explore landscape response to tectonics under a variety of different climate scenarios. We explore different rates of uniform rock uplift, and tilting scenarios with rock uplift rate either increasing or decreasing linearly along the profile. Climatic forcing is based on temperature, either uniform temperature or simplified 100kyr glacial cycles. The initial condition is a concave profile, typical of a fluvial landscape, with a short plateau reach at its head that acts as a source of windblown snow. In all circumstances a uniform climate results in a steady decrease in glacier size, as found in previous studies. Both uniform and variable temperatures result in headwall lengthening, but the valley floor response is markedly different. Uniform temperature efficiently creates a cirque form, with a shallow downvalley gradient. Under warmer conditions, the cirque glacier only lowers the upper part of the profile, whereas the valley glacier that forms under cooler conditions flattens most of the valley floor. Variable temperature causes erosion along a greater proportion of the valley floor for a given mean temperature, but less dramatic decreases in downvalley gradient. Changes in the imposed tectonic regime are weakly reflected in the valley profile, but strongly influence headwall relief. These results, and their consistency with field observations, emphasise the potential for substantial landscape modifications by glaciers that are strongly dependent on the details of both tectonic and climatic forcing.
H53C-1381
The Glacial Buzzsaw in the Northern Basin and Range: the Importance of Glacier Size and Uplift Rates
The role of glaciers in limiting mountain range elevations is an important component of studies linking tectonic uplift and climate-driven erosion. Recent investigations suggesting that a glacial buzzsaw effect can efficiently offset rock uplift in tectonically active settings have concentrated on regions that have held large glaciers (10s km long at Last Glacial Maximum, LGM). However, little work has addressed the role small glaciers may play in controlling range topography. This study looks at the effectiveness of smaller (<10 km) glaciers at limiting peak and ridge elevations in both slow and relatively rapid rock uplift settings. The Lost River and Lemhi Ranges, Idaho, and the Beaverhead-Bitterroot Mountains, Idaho-Montana all experience slow rock uplift, with slip rates <0.3 mm/yr on the range-bounding normal faults. Here, swath-elevation profiles show that maximum elevations correlate well to estimates of both LGM and mean Quaternary equilibrium line altitudes (ELAs). Furthermore, peaks in hypsometry and minima in slope-elevation profiles correspond to ELAs, suggesting that small glaciers can efficiently limit range elevations where rock uplift is slow. The Teton Range, Wyoming, experiences 5-10 times faster rock uplift. In general, elevations, slope profiles, and hypsometry all correlate to both LGM and mean Quaternary ELA estimates, although supra-elevated peaks do penetrate through this zone. Comparisons of valley long-profiles show that glacier size is important in controlling valley form under more rapid rock uplift. Small (<5km) glacial valleys perched high on the range front have profiles that have steepened in response to the rapid rock uplift. In contrast, larger (>8km) valleys extend back beyond the high peaks of the range front, and have housed glaciers that have eroded deep into the range, maintaining shallow gradients. Feedback mechanisms are important in snow accumulation on the larger glaciers, which receive extra inputs of snow from the neighbouring high peaks. Glacier size is apparently key in controlling a glacier's ability to keep pace with rock uplift.
H53C-1382
Long-Term Glacial Erosion in the Coast Mountains, British Columbia, Canada from Low- Temperature Thermochronology
We integrate a dense suite of low-temperature thermochronometry samples with numerical models to constrain long-term (>106 yr) denudation within the heavily glaciated Coast Mountains, British Columbia, Canada. We synthesize 84 new and published apatite (U-Th)/He (AHe), 13 apatite fission track (AFT), 9 zircon (U-Th)/He (ZHe) and 6 zircon fission track (ZFT) cooling ages that range from 1.4 - 15.4 Ma (AHe), 5.2 - 34.5 Ma (AFT) 2.4 - 26.6 Ma (ZHe) and 18.6 - 55.2 Ma (ZFT). This span of ages allows us to quantify both the pre- and post-glacial history of the region, as well as detailed spatial variations in erosion. Samples were collected over a ~2500 km2 region, spanning 4 km of relief across glaciated valleys and ridges. Cooling ages generally increase in age with increasing sample elevation throughout the region. However, distinctly different linear age-elevation relationships exist above and below ~2000 m. As a consequence of this, ages above 2000 m average 7.5, 18.0, 25.0 and 45.4 Ma for AHe, AFT, ZHe and ZFT, respectively, while below 2000 m the averages decrease to 4.2, 10.7 15.7 and 39.8 Ma. We use the spatial distribution of cooling ages and erosion rates calculated from a 1-D thermo-kinematic numerical model to quantify the effect of glaciation on long-term erosion rates. Calculated rates range from 0.2 - 2.2 mm/yr over the last 55 myr. Relatively young cooling ages and higher erosion rates occur preferentially at low elevations indicating increased glacial intensity in the lower flanks of the broad U- shaped valleys. Best-fit regression lines and inverse modeling of samples along vertical transects suggest most rapid cooling is constrained within the last ~6 Ma, coincident with the onset of regional alpine glaciation. Our results are consistent with denudation of a former paleotopographic high offset ~16 km to the SW from its present position. Current work in progress is developing and applying a 3-D numerical model to better constrain the spatial and temporal variations in erosion rates.
H53C-1383
Vertically mixed and unmixed: Do surface features tell the whole story? An investigation of glacial regolith profiles using in-situ produced cosmogenic radionuclides
Whether a regolith is unmixed or mixed is critical to determining its erosion rate or age from in situ-produced cosmogenic nuclides. We use in situ-produced 10Be and 26Al in quartzite clasts extracted from depth profiles to investigate mixing of a periglacially-sorted till blanketing a plateau in the northern Swedish mountains. Our data indicate significant intra-site variations from a completely unmixed to a fully mixed regolith. We conclude that caution must be exercised in assuming that an entire regolith is either unmixed or mixed from interspersed depth profiles and that the degree of mixing may differ significantly from that indicated by observation of surface features. From the difference between the surface isotope concentration of an unmixed profile and the average isotope concentration of a fully mixed profile, we confirm that the regolith is a glacial till and that it could have been emplaced in a single event. Incorporating isotope concentrations, 26Al/10Be ratios, and an isostasy and ice sheet burial model we date the till emplacement to the Saalian glaciation (~ 200 to 130 ka).
H53C-1384
A Reevaluation of the DeKalb Mounds of Northern Illinois
We re-examined the classic DeKalb Mounds of Northern Illinois in order to better understand their genesis, morphology, spatial distribution, and usefulness in reconstructing deglacial and postglacial environments. Flemal et al. (1973) interpreted the mounds to be relict pingos resulting from an intense periglacial environment during the late Wisconsin. Thousands of mounds occur in the study area. The mounds range in size from 20 m to more than 5 km in diameter and typically rise 1 to 8 m above the surrounding loess mantled till surface. These low relief mounds are composed of either a raised border surrounding a low center ("donuts") or are flat-topped ("pancakes"). Modern soil properties are strongly related to mound position and subtle sedimentologic and topographic changes. In some locations multiple mounds appear to be superimposed on one another. The typical mound stratigraphy includes subglacial diamicton at the base, less than 1 m of glaciofluvial sands and gravels or debris flow diamicton, 1 to 6 m of fossiliferous, rhythmically bedded lake sediment, 0.3 to 3 m of glaciofluvial sands and gravels or debris flow diamicton, all capped by approximately 1.2 m of loess. The lake sediment is rich in ostracodes and tundra plant remains. Smaller mounds are typically symmetrical while the larger mounds are more elliptical in shape with a long axis trending northeast to southwest. Many of the larger mounds appear to be aggregates of smaller lakes that coalesced as glacial ice stagnated. Our reinterpretation of the mounds is that there genesis is related to deglaciation and ice-stagnation and not to post glacial permafrost processes. We interpret all of the DeKalb Mounds to be ice-walled lakes and not pingos.
H53C-1385
Contemporary proglacial aeolian sediment transport in West Greenland
Glacial erosion processes produce significance quantities of fine sediments that are washed out from beneath glaciers by meltwater. When deposited on the glacier floodplain they dessicate and strong ice-driven winds can entrain and transport them across the landscape resulting in the formation of sand dunes and loess, and adding very fine particles (dust) to the atmosphere. Recent studies suggest that locally-generated dust can play an important role in regulating albedo and the melting rate of glaciers. Very few field process studies have examined the relationship between sediment-delivery to the proglacial floodplain by meltwater and the subsequent aeolian erosion and deposition of these fine sediments. This research reports the use of semi-isokinetic directional sediment samplers to make an initial assessment of the rates of transport of dust and sand in Sandflugtdalen, a valley adjacent to the West Greenland ice sheet. Vertical arrays (z(m) = 0.18, 0.43, 0.85, 1.4) of samplers were deployed in a down valley transect over a distance of 4 km. Trapped sediments were retrieved after intervals of 1 week and 9 weeks. The mass of sediment collected in the traps varied from 0.002-3.62 g cm2 wk-1. As expected, near surface traps collected more, and coarser, sediment than those deployed at 1.4 m height but the decrease in mass of sediment with height was highly variable. The array closest to the glacier trapped the greatest quantity of suspended sediment and the density of suspended sediment decreased with distance down valley. The flux of aeolian sediment comprises clays, silts and sand-sized particles. Areas of aeolian entrainment, transport and deposition are closely linked to the development and distribution of sediments on the proglacial floodplain which varies considerably in terms of surface roughness. At the east end of the valley, close to the ice sheet, aeolian sediment flux is controlled by sediment supply and topography rather than wind speed. Further down valley, recycling of sediments by fluvial and aeolian activity is important and wind speed is an important controlling factor. Within the dunefields in the valley, surface roughness is determined by topography and also by vegetation. Maximum aeolian sediment transport is in early summer before the annual vegetation cover within the dunefield has fully developed.
H53C-1386
Comparing the Effectiveness of Ground-Penetrating Radar in Imaging Siliciclastic And Mafic- Volcaniclastic Dune Sands
Experiments using ground-penetrating radar (GPR) systems in two, different dune sediment environments allow comparisons of the relative effectiveness of subsurface imaging and feature detection. One experiment was carried out in the Coral Pink Sand Dunes (CPSD) in southern Utah, and a second in the Grand County Off Road Vehicle area in Moses Lake (ML), Washington. Both experiments used a MALA GPR system with 500MHz antenna and similar data sampling and acquisition parameters. The dunes at the CPSD site are comprised of nearly pure, very well sorted quartz sands. These sharply contrast with dunes at the ML site which are comprised of basalt-rich (up to 80%) sands. The ML site was selected as a terrestrial analog to Martian dunes that have been shown in other studies to have a similar mineralogy. As with other quartz dune studies, radar images gathered at the CPSD site clearly show cross-bedding structures and were able to identify the bedrock/dune interface as well as the locally shallow water table. The imagery collected at the ML site was not as clear, but some dune structures, ash beds, and water are visible in the imagery. We propose that thee higher basalt content at the ML sites results in greater signal loss than in the siliciclastic sands at the CPSD site. The reduced signal transmissivity in the mafic sands may have implications for selection of GPR instrumentation in future Mars investigations.
H53C-1387
Determining soil erosion rate and sediment residence time on hillslopes using radionuclides in semiarid grassland of Mongolia
Radionuclide technique has been attracted increasing attention as an alternative method for soil erosion studies. Recently, the potential use of 210Pb which is a naturally occurring radionuclide has been explored as an alternative for 137Cs. This study estimated long-term soil erosion rate by using 137Cs and 210Pbex in semiarid grassland of Mongolia; where overgrazing has been considered to be the cause of intensive soil erosion and land degradation. A couple of the experimental watershed (<10 ha) was established in areas of different grazing conditions. In order to determine the spatial distribution of radionuclides, more than thirty soil cores were collected within the watersheds. The activity of radionuclides was measured using HP N-type Gamma ray detector in the laboratory of University of Tsukuba. The gain/reduction rate of radionuclides to the reference inventory was converted to soil erosion rate by using the Diffusion and Migration model (He and Walling, 1993). Based on 137Cs inventories, an intensive erosion was detected in the overgrazed watershed; whereas soil erosion was negligible at the watershed where there was currently no overgrazing. In contrast, 210Pbex inventories exhibited that the soil erosion rate was very much smaller than that based on 137Cs inventories; especially at the no overgrazed site. In addition to this, along a transect on erosion slopes, the 137Cs inventory tended to decreased downward, while the 210Pbex/137Cs inventory ratio and the activity of surface soil (< 20 mm) of 210Pbex increased. The downslope increase of 210Pbex activity may be attributed to the direct deposition of atmospheric 210Pbex onto soil surface. Therefore, given the direct deposition rate of 210Pbex, the sediment transport on hillslopes should be determined from the increasing rate of 210Pbex in the surface soil. A simple budget calculation of 210Pbex activity of surface soil roughly estimated the residence time of sediment on hillslopes. The results of this study suggest that, under the environments where the sediment transport is very slow, 210Pbex is seen to be inappropriate for the estimation of soil erosion rate because the conversion models for the estimation of soil erosion rate assumes eroded sediment to be readily removed from hillslope; instead, the use of 210Pbex should be proposed to quantifying the residence time of sediment on hillslopes.
H53C-1388
A New Approach for Estimating Background Rates of Erosion Using Concentration of Meteoric 10-Be Adhered to River Sediment: Application to the Rapidly Eroding Waipaoa Basin, New Zealand
New and existing data suggest that the concentration of atmospherically- produced, meteoric 10-Be adhered to river sediment provides a proxy for basin-scale erosion rates. Although the widely applied method of analyzing in situ produced 10-Be in river sediments has proven useful for estimating pre-anthropogenic rates of erosion in a variety of environments, there are lithologic limitation. In contrast, measuring the concentration of meteoric 10-Be adhered to river sediment allows erosion rate analysis in landscapes underlain by quartz-deficient or fine-grained lithologies, as well as in basins where the concentration of quartz varies spatially. By assuming that basins are in an overall isotopic steady-state, that erosion is rapid enough that decay is negligible, and that the integrated delivery rate of 10-Be from the atmosphere (D10-Be) can be estimated, basin-scale mass loss rates (Ms) can be solved by equating the 10-Be flux in from the atmosphere with the flux of 10-Be out of the basin on sediment (C10-Be) and expressed as sediment yield per unit area (Ys). Fin = Fout D10-Be * A = Ms * C10-Be Ms = (D10-Be * A)/ C10-Be Ys = D10-Be / C10-Be To validate this new approach, we examined the limited data that do exist and found reasonable correspondence between erosion rates estimated from meteoric 10-Be concentrations and estimated by other means. As a first application, we use meteoric 10-Be in river sediment to estimate basin-scale erosion rates from catchments within and near the mud-stone dominated Waipaoa River Basin draining the tectonically active east coast of New Zealand's North Island. Near total conversion of indigenous forest to pasture over the past century in the Waipaoa Basin has resulted in some of the most dramatic and widespread erosional features on the planet, and contemporary sediment yields that rank among the highest in the world (~7 million kg/(km2 * yr)). The amount of meteoric 10-Be adhered to eight river sediment samples suggests that modern-day sediment yields are at least seven times higher than natural rates of sediment generation. This finding is in tight agreement with other estimates of pre-settlement sediment discharge from the Waipaoa Basin derived from middle shelf and nearby lake cores (Kettner et al., 2007; Page and Trustrum, 1997). Tributary basins (n=4) draining portions of the Waipaoa Basin dominated by landsliding in shallow soils yield an average background sediment generation rate of 106 ± 105 kg/(km2 * yr), assuming a deposition rate of 1.3 million atoms 10-Be/(cm2 * yr). Conversely, sediment shed from a basin dominated by severe gullying contains ~four times less 10-Be due to the contribution of deeply sourced material containing little or no meteoric 10-Be. Large basins to the north and south of the Waipaoa (n=3) yield similar background rates of sediment generation ranging from 0.25 to 1.6 million kg/(km2 * yr). Meteoric analysis of an additional 40 samples, as well as cross-calibration between in situ produced and meteoric 10-Be in 19 quartz-bearing samples will further test the robustness of this new approach for estimating natural rates of sediment generation and erosion.
H53C-1389
Terrace ages, timing of colluvial hollow denudation, and sediment routing in Pancho Rico Valley, Coast Ranges of central California (Monterey County), USA
Field observations and mapping, combined with optically stimulated luminescence (OSL) dating reveal the effect of the Pleistocene-Holocene transition on colluvial hollows and stream channels in Pancho Rico Valley (part of the Pancho Rico Creek [PRC] drainage basin). Three samples of PRC fill-terrace sediment yielded OSL ages (14.1 +/- 1.41 ka, 13.8 +/- 1.59 ka, 13.1 +/- 1.30 ka) that are coincident with the Pleistocene-Holocene transition. These PRC fill terraces, which overlie straths, together with PRC-tributary fans and PRC-tributary valley fills comprise a physically continuous geomorphic surface and are collectively mapped as Qta. In the context of the terrace ages, colluvial hollow morphostratigraphy and debris flow deposits within Qta sequences indicate sediment was excavated from hollows and delivered to PRC via slope failures. The slope failures were triggered by relatively high frequency, potent storms during the Pleistocene-Holocene transition. Landslides caused by the stormy climate of this climatic transition may have also led to creation of new hollows in incompetent bedrock. Hollows were completely stripped of sediment sometime during (or after) the Pleistocene-Holocene transition, and drier Holocene climatic conditions have precluded refilling of older hollows or infilling of newly created hollows. Vegetation on Pancho Rico Valley hillslopes is relatively sparse because of relatively low precipitation during Holocene-time, and the predominance of south-facing slopes in the strongly asymmetric Pancho Rico Valley. Holocene-time refilling of colluvial hollows has not been established because of scouring caused by slope-wash erosion on sparsely vegetated hillslopes. Qta morphostratigraphy provides insight regarding Late Quaternary sediment routing and geomorphic processes in Pancho Rico Valley. The absence of Qta lobes in the alluvial fan that exists where PRC flows into the Salinas Valley indicates most sediment discharged from Pancho Rico Valley colluvial hollows is stored in Qta PRC terraces, Qta tributary fans, and Qta valley fills. Qta treads are base level for many Pancho Rico Valley hillslopes, and because Qta is mostly intact, Pancho Rico Valley hillslopes have been isolated from base-level fall caused by ongoing Holocene-time PRC incision. Debris flows like those impinging on the PRC channel during Qta time have been uncommon in the Holocene because there is insufficient sediment within colluvial hollows for slope failures to occur. Instead, sediment generated on hillslopes and mobilized by slope wash is deposited on Qta treads.
H53C-1390
Late Holocene Soil Stratigraphy and Geochronology of Alluvial Sedimentation in the Sonoran Desert, Arizona
The integration of soil stratigraphic investigations and radiocarbon dating at two sites in combination with geomorphic mapping at scales of 1:50k and 1:5k offer insight to the timing and magnitude of alluvial sedimentation during the late Holocene within the Sonoran Desert near Yuma, Arizona. Mapping at 1:50k was performed over an area of 3400 km2 and alluvial landforms were labeled Qf1 to Qf5, from oldest to youngest, using 1- and 5-meter resolution satellite imagery within the U.S. Army Yuma Proving Ground (YPG). Approximately 70% of the identified landforms within YPG are Quaternary alluvial fans, alluvial plains, and active washes, whereas the other 30% consist of mountain highlands, pediments, and badlands. In the southwest portion of YPG near Muggins Mountains, alluvial fan terraces (Qf4) positioned 0.5 m above active washes are characterized as having moderate bar-and-swale microtopography, moderately developed desert pavement, and a Av/ Bw/ Cky/ Cky1/ Cky2/ Cky3 gravelly soil profile. A large piece of charcoalized Ironwood ( Olneya tesota) was recovered from a depth of 0.75 m and yielded three AMS 14C dates that range from 3330 to 2860 cal yr B.P. Geomorphic mapping at a scale of 1:5k indicates that in an area of 25 km2 at the site, the distribution of late Holocene alluvial fan terraces comprise 17% of the surrounding Quaternary alluvium. Similar aged alluvial features were observed about 70 km to the north near South Trigo Peak at YPG. Terraces of a broad and flat alluvial plain positioned 0.5 m above active channels are characterized as having moderate bar- and-swale microtopography, poorly developed desert pavement, and a AC/ C/ Bwkb1/ Bwk1b2/ Bwk2b2/ BCkb2/ Bwkb3 sandy soil profile. A terrestrial gastropod shell ( Lymnea sp.) fragment was recovered from a depth of 0.5 m and yielded an AMS 14C date of 2360-2310 cal yr B.P. Additional geomorphic mapping at 1:5k shows that in an area of 30 km2, the distribution of late Holocene alluvial plain terraces and equivalent alluvial fans include 35% of the surrounding Quaternary alluvium. Of the total Quaternary alluvium identified at 1:50k, 10% of the landforms consist of gravelly alluvial fans and sandy alluvial plains deposited between 3330 and 2310 cal yr B.P. The numerical ages and associated soil development provide evidence of regional alluvial sedimentation near Yuma, Arizona, and provides well constrained geomorphic data for paleoclimatic modeling in the Sonoran Desert during the late Holocene.
H53C-1391
Topographic Analysis of Landscape Morphology and Vegetation Patterns in a Semiarid Basin in Central New Mexico
We examine the linkages between basin morphology, vegetation patterns, and lithology in a semiarid basin located at the upper portions of the Rio Salado River in westcentral New Mexico, USA. In our study area, basin elevation plays a dominant role in the distribution of vegetation types from shrublands in lower elevations to grasslands and forests as elevation increases. Arguably the observed vegetation cover of the basin reflects the vegetation types of the modern climatic conditions of the late Holocene. This region is believed to be covered by spruce-fir, mixed-conifer, and subalpine forests prior to Pleistocene-Holocene climatic transition. We analyze mathematical properties of landforms including slope-area scaling relations, power-law distributions of areas, probability distributions of slopes, and drainage densities of more than 20 subcatchments with varying vegetation types. We found statistically significant differences in the organizational of properties of landforms as related to different vegetation types. These findings suggest a rapid geomorphologic response in the basin with climate- induced vegetation shifts that had occurred in the Holocene.
H53C-1392
Rainsplash as an Advection-Dispersion Process, With Implications for Plant-Soil Interactions
Raindrops falling on loose sediment transfer part of their momentum to the sediment grains, which are ejected radially outward from the impact site. This process, referred to as rainsplash, plays an important role in sediment transport and landscape evolution. Describing rainsplash within the framework of the Master Equation and the Fokker-Planck Equation suggests that the transport of grains is an advection-dispersion process. By taking into account the intermittency of grain motions activated by raindrop impacts, the formulation indicates that gradients in raindrop intensity, and thus grain activity, can be as important as gradients in grain concentration and surface slope in effecting transport. Outdoor lab tests simulating rainfall over loose sand, on both flat and sloped surfaces, provide evidence to support this advection-dispersion model of rainsplash. Furthermore, sheltering a portion of the test surface from rain causes differential rainsplash to occur. As a result, sediment builds up as a mound beneath the shelter. The behavior of the shelter imitates that of plant canopies, which suggests that rainsplash contributes to the development of sediment mounds beneath shrubs in desert environments. These conclusions are consistent with observations of natural mounds at the Sevilleta Long Term Ecological Research station and Bandelier National Monument in New Mexico.
H53C-1393
The effect of slope angle on splash detachment in steep forest plantation
To study splash detachment rate and investigate the effects of rainfall and slope angle on splash detachment, the field observation of splash detachment was conducted for five months using 27 splash cups under natural rainfall events in Japanese cypress (Hinoki; Chamaecyparis obtusa) plantation in the Shimanto River watershed, southern Japan. In this plantation forest, the unit kinetic energy of throughfall (unit KE; J/m2/mm) was found to be constant independent of rainfall intensity. The total rainfall over six observation periods was 853 mm and the maximum rainfall intensity for 1 h ( RI1h) varied from 8.0 to 19.6 mm h-1. A significantly high coefficient of linear regression was found between RI1h and the average splash detachment of all splash cups over six periods, although the splash detachment from the individual cups had larger variations with RI1h. This variation in splash detachment may attribute to the spatial variability in soil surface condition such as slope angle. In the relationship between the splash detachment and slope angle, no correlation was found over the entire periods. However, different correlations were found among the observation periods due to the differences in rainfall intensity. The splash detachment from a lower slope angle (14°) exhibited a strong relation with the maximum rainfall intensity for a shorter period, such as 10 to 30 minutes. In contrast, the splash detachment from a slope angle of over 35° exhibited high correlation with the maximum rainfall intensity for 3 h, suggesting that longer time is required for ponding in steeper slopes than gentler slopes. In gentler slopes, prolonged rainfall may cause the higher ponding depth, resulting in reducing the raindrop impact and less splash detachment. Therefore, under the forest canopies, the effect of slope angle on the rainfall parameter should be incorporated into the future splash erosion model.
H53C-1394
Influence of canopy thickness on throughfall amount and kinetic energy under different canopy saturation conditions: an indoor experiment with a Japanese cypress (Chamaecyparis obtusa) stand
To assess the influence of canopy thickness on throughfall amount and kinetic energy under different canopy saturation conditions, indoor laboratory experiments were conducted involving water sprinklers and a transplanted tree stand of Japanese cypress (Chamaecyparis obtusa) of 9.8 m high. Artificial rainfall was applied onto the tree for 15 min with an intensity of 39.8 mm h-1. Throughfall amount and raindrops measured at eight points for four canopy structures generated by staged branch pruning, each first branch height was 2, 3, 4, and 5 m, respectively. With the thinning of the canopy thickness, 1) initial throughfall amount increased depending on the decrease of canopy storage, 2) the abundance ratio of large drops generated as the drips increased owing to the decrease of splash droplets generated by the impact of the drips onto the foliage in the lower canopy layers, 3) the drips with higher velocities generated from the upper canopy layers increased, and consequently, 4) throughfall kinetic energy increased. The canopy thickness has an important effect on the variability of throughfall amount and kinetic energy due to the change of processes of canopy saturation and the drip generation in the canopies.
H53C-1395
The effect of surface cover on infiltration rate in steep forest plantations
The Japanese cypress (Hinoki; Chamaecyparis obtusa) is a major commercial tree species in Japan, and without thinning of high-density stands, canopy closure prevents development of understory vegetation. Therefore there is a concern for overlandflow and sediment yield due to infiltration rate lowering from steep hillslopes of Japanese cypress plantation. We developed a light-weight rainfall simulator based on the design of Meyer and Harmon (1979). A flat fan Veejet 80150 spraying nozzle (Spraying systems Co., USA) is mounted on the manifold at 2.13 m high from the plot surface. The nozzle oscillates so that the spray fans swept across the targeting 1m x 1m plot. The Veejet 80150 spraying nozzle produces large raindrops larger than 2 mm in diameter, and can simulate the high raindrop kinetic energy of natural storm. A targeted rainfall rate is 180 mm/h. Total 25 sprinkling experiments have been conducted on 35-degree hillslopes with varying surface cover. We obtained the minimum infiltration rate of 14 mm/h where the surface cover is very little. The infiltration rates were plotted against the total understory vegetation and dry weight of total surface cover including litter. The infiltration rate increased with the increasing total surface cover, and higher regression coefficient is obtained for the case of the total surface cover. These results will contribute to the future modeling studies of overlandflow occurrences for the catchment scales.
H53C-1396
Analysis of Runoff Generation under Different Canopies in Mountainous Slope
In order to examine the runoff generation processes in mountainous slope under Larix leptolepis (S1)and Abies holophylla (S4) two small plots were monitored. The rainfall and throughfall of both sites showed respectively a good correlations, whereas the quantity and intensity of throughfall in S4 exhibited larger than those in S1 and the mean lag time for throughfall generation from rainfall start in S4 was shorter than that in S1. The data of throughfall and runoff exhibited that the runoff discharge and runoff peak in S4 were smaller than those in S1, whereas the mean lag time for runoff generation from throughfall start in S4 was longer than that in S1. The total lag times from rain start to runoff generation of S1 and S4 showed similar, whereas each the lag time for throughfall and runoff generation exhibited different with sites. The results of this study suggest that the influences of different canopy species effects on throughfall characters and surface cover conditions, especially the lag time for throughfall and runoff generation, and which were distinctly reduced by the intensity increasing of rainfall and throughfall during the lag time.
H53C-1397
Field Measurements of Macroflow and Macroporosity on a Hillslope
Even though the impact of macropore on the hydrological process at the hillslope scale has been widely recognized, developments and distribution of macropore have not been investigated in conjunction with the characteristics of the hillslope i.e. topography, soil property and bioactivity. In this study, macropore properties, such as macropore flow and saturation hydraulic conductivity were measured for a hillslope in Gwangneung Research Forest, located northern South Korea. An intensive field survey provided a refined Digital Elevation Model (DEM) for surface and subsurface topography. Spatial distributions of upslope area and topographic wetness index can be obtained through the digital terrain analysis. The selected points, 19 monitoring locations, were distributed along transacts of the digital contour map. Vertical fluxes through macropores were measured using a tension infiltrometer at the depth of 10 cm from the surface. Spatial and temporal distributions of soil moisture were obtained using the on-line measurement system, TRASE, previously installed in the study area. Soil moistures for the aforementioned points were measured at 10cm and 30cm depths from the surface. The results from tension infiltrometer experiments indicate that macropore flows were ranged between 21% and 94% and Measured macroporosities were varied from 1.4% to 47%. Macropore flow and macroporosities tend to increase in the downslope direction. Analyses of seasonal soil moisture variation also explain the distribution of macropores. The increasing tendency toward the outlet can be explained by the subsurface erosion, surface and bedrock topography and bioactivities.
H53C-1398
Quantifying the effects of soil compaction on runoff and erosion
Anthropogenic activities, such as construction, alter the Earth's surface by disturbing the landscape and intensifying erosion processes. As construction continues rapidly worldwide, emphasis is being focused on improving construction techniques for minimizing soil erosion and improving stormwater quality. The challenge on construction sites is that soil surface conditions are dramatically modified in terms of depth and density throughout the various phases of earthwork. In addition, the current understanding of how altered soil density impacts infiltration and soil erosion during rainfall events is limited. The objective of this research is to quantify and explain the effects of soil compaction on runoff and erosion for bare soil conditions. Laboratory experiments were performed in the Soil Erosion Research Laboratory (SERL) at San Diego State University. SERL experiments utilize a 3-m by 10-m tilting soil bed with a soil depth of 0.5 meters and a slope of 33 percent. The simulated storm event consists of two periods: 51 mm/hr (2 in/hr) for 20 minutes, followed by 102 mm/hr (4 in/hr) for 30 minutes. The storm event is based on an ASTM design storm intended to simulate BMP failures. All experiments where preformed using a Sandy Loam soil (USDA). Soil densities were varied from 1.2 to 1.6 g/cm3. Based on the experimental data sediment to runoff ratios vary from 0.3 to 0.5 Kg/L depending on soil density. Preliminary results show that the relationship between soil erosion and soil density initially decreases to a minimum, then increases until becoming stationary. This relationship is explained by characterizing the runoff and erosive processes at varying soil densities. A primary factor in this research is the decrease infiltration capacity with increasing soil density. http://spatialhydro.sdsu.edu
H53C-1399
Performance Probability Distributions for Sediment Control Best Management Practices
Controlling soil erosion and sediment transport can be a significant challenge during the construction process due to the extent and conditions of bare, disturbed soils. Best Management Practices (BMPs) are used as the framework for the design of sediment discharge prevention systems in stormwater pollution prevention plans which are typically required for construction sites. This research focuses on commonly-used BMP systems for perimeter control of sediment export: silt fences and fiber rolls. Although these systems are widely used, the physical and engineering parameters describing their performance are not well understood. Performance expectations are based on manufacturer results, but due to the dynamic conditions that exist on a construction site performance expectations are not always achievable in the field. Based on experimental results product performance is shown to be highly variable. Experiments using the same installation procedures show inconsistent sediment removal performances ranging from (>)85 percent to zero. The goal of this research is to improve the determination of off-site sediment yield based on probabilistic performance results of perimeter control BMPs. BMPs are evaluated in the Soil Erosion Research Laboratory (SERL) in the Civil and Environmental Engineering department at San Diego State University. SERL experiments are performed on a 3-m by 10-m tilting soil bed with a soil depth of 0.5 meters and a slope of 33 percent. The simulated storm event consists of 17 mm/hr for 20 minutes followed by 51 mm/hr for 30 minutes. The storm event is based on an ASTM design storm intended to simulate BMP failures. BMP performance is assessed based on experiments where BMPs are installed per manufacture specifications, less than optimal installations, and no treatment conditions. Preliminary results from 30 experiments are presented and used to develop probability distributions for BMP sediment removal efficiencies. The results are then combined with spatial and temporal distributions of perimeter sediment loadings for a construction site to estimate the time dependent risk of off-site sediment discharge over the duration of a project (ex., 0, 25, 50, 75 and 100 percent complete). The results are used to highlight the importance of considering all phases of construction when developing stormwater pollution prevention plans. http://spatialhydro.sdsu.edu
H53C-1400
Quantifying the spatial and temporal distribution of sediment loading to the perimeter of construction sites
Large scale construction projects can dramatically change the drainage pattern of a landscape. During the earthwork phases of construction internal drainage networks are altered even though connections to the existing networks beyond the site boundary must be maintained. These modifications result in large differences in contributing area and sediment loading to specific points along the site perimeter. While most construction projects require stormwater pollution prevention plans to control runoff, these plans often only consider pre- and post- site conditions. Due to significant increases in soil erosion and off-site sediment discharge potential during construction, a better understanding of the sediment loadings throughout the construction process is needed. This study investigates the spatial and temporal distribution of drainage characteristics and sediment loading to the perimeter of construction sites using GIS-based methodologies A case study is presented to illustrate the approach for initial, 25, 50, 75 and final grading. The study site has a land area of 210 ha and is located in southern California. The site has a perimeter length of 3.6 km, and using a 0.6 meter elevation grid, the perimeter consists of 7191 pixels, with each pixel representing one watershed draining to the perimeter. The results show how the distribution of drainages, flow lengths and estimated event sediment yield change over time. The maximum drainage area to site boundary for pre- and post- conditions is 68 and 39 ha, respectively. Predicted sediment yield to the perimeter of the site for pre- and post-grading conditions is 12 and 20 t/ha, respectively. The maximum sediment yield from an individual watershed (i.e., point loading to the perimeter of the site) decreased from 230 to 30 t/ha for pre- and post-grading conditions. Based on the grading schedule, the site is characterized at five time intervals. Preliminary results suggest that quantifying the sediment loading to the site perimeter at only the initial and final stages of grading can underestimate the potential maximum sediment load to the perimeter of the site. http://spatialhydro.sdsu.edu
H53C-1401
Root Throw and Sediment Transport in the Rocky Mountains of Western Canada: Field and Modelling Investigations
Sediment transport associated with root throw was investigated in Kootenay National Park in southeastern British Columbia. Tree toppling and root throw may result in sediment transport, as soil associated with the root wad is upheaved in the form of a root plate, which eventually disintegrates and deposits on the ground forming a mound. Although root throw is a widely recognized process, its role as an agent of sediment transport has not been widely considered. This study provides critical field data documenting the role of episodic root throw on sediment transport at a local scale (order of magnitude approximately 1 m). We then use these data in conjunction with a forest population dynamics model to consider the contribution of root throw to larger-scale sediment routing and its temporal dynamics. A detailed field measurement program documented characteristics associated with root throw in burned and pre- fire scenarios (e.g., root plate dimensions, direction of tree fall, root wad disintegration). Sediment transport rates due to root throw are relatively low; approximately two orders-of-magnitude lower than typical creep rates defined in the literature. However, in a landscape where larger mass movements are relatively uncommon, this small transport plays a role in sediment transfers, and contributes to soil mixing and formation of microtopography. In the post-fire scenario, an increase in root throw occurred as fire-killed trees toppled in the first years following the fire. The removal of vegetation also exposed root plates to erosion, and disintegration rates increased. A tree population model is developed to simulate tree recruitment, growth, mortality, and toppling rates over time scales greater than 1000 years. In the model, fire events occur as a stochastic disturbance which kills all trees, and recruits new trees. Thus, the model cycles through forest generations with lifespan determined by fire events. The number of trees at any time interval is the net result of recruits minus the proportion that die within the time interval by either wildfire or other ecological causes e.g., competition. Tree dbh is based on tree age and age- related growth rate, with a minimum dbh required before a tree will uproot significant amounts of sediment. Root plate volume, width, height are based on dbh and calibrated using the field data. Some literature suggests that dead trees uproot significantly less sediment than live trees when they fall over, and this option can be incorporated in the model. Subsequent disintegration rates of root plates are calibrated using the field data to derive transport rates in the model. The temporal sequencing of sediment transport rates due to tree topple over millennial time scales is influenced by the tree population dynamics and the associated fire return intervals. Results of the simulations show that peaks in sediment transport occur at approximately one half and one forth of the average fire interval. The sediment transport becomes more continuous and less pulsed as the fire interval increases. Both of these effects are a result of the interaction of tree topple with fire and competition and old age.
H53C-1402
The Abnormal Pore Pressure Response Between the Interface of Colluvial and Bed Rock in a Landslide Area
The Hungtsaiping Landslide triggered by Chi-Chi earthquake covers an area of hundreds of hectors. The rock mass is covered by a thick colluvial that is as deep as 82m. To explore the sliding mechanisms, GPS, inclinometer and fiber optic displacement and water pressure sensing systems were used to monitor the status of the landslide. Among them, three piezometers installed in the interface of colluvial and bed rock automatically recorded lots of valuable data during July of 2005 to the end of 2006. Based on records of the pore pressure varied with rainfall, normal and abnormal responses were identified and discussed. The seasonal variation of pore pressure was less than 30 kPa. During the period of heavy rainfall, the variation of pore pressure was less than 5 kPa. Besides the normal response of pore pressure increased quickly after rainfall, abnormal response was observed which the pore pressure decreased during the period of heavy rainfall. In the same period of the pore pressure dropped suddenly, the surface and subsurface displacement was also observed. Accordingly, the decreasing of pore pressure could original from the dilation in the sliding surface which induced a negative excess pore pressure. The negative excess pore pressure occurs when the rainfall intensity greater than 30mm/hr, the accumulated precipitation greater than 124mm and the duration of rainfall greater than 15 hr.
H53C-1403
Granular Flows: A Discrete Look at Particle-Bed Interactions
It has been argued that debris flows play a fundamental role in shaping steep topography, yet questions remain about the magnitude of debris flow induced landscape change and the processes by which they erode. If the processes that control the rate at which debris flows and related mass movements cut steep bedrock channels were better understood, prediction of their impact on steepland evolution could be possible. An important control on the erosion of steep bedrock channels is the amount of stress transferred to the bed from the overlying flow. Field evidence from many investigations indicates that impacts from discrete particles could be a major contributor to bedrock erosion, as could particle sliding. To investigate the interaction between a granular flow and the subjacent bed we focus our investigations on three characteristics of a dry granular flow: type of particle- bed interaction (i.e. purely collisional with no slip vs. collisional with slip), dominant flow regime (quasi-static to rapid), and changes in basal stress with measurable flow attributes. We explore these flow characteristics using a Lagrangian discrete element model originally developed by Cundall and Strack (1979). With this type of computer simulation, particle-particle and particle-boundary interactions are modeled explicitly for every particle and boundary in the system. The model is first validated with comparisons to: analytic solutions for single particle impacts, energy conservation for a perfectly elastic system, and documented analog experiments. The model is then used to calculate the relative importance of sliding vs. non-sliding contacts. For each sliding contact the average distance traveled while in contact with the bed is recorded. The dominant flow regime is determined by observing whether force is transferred through enduring contacts and stress chains or via short-lived particle- particle contacts. By identifying the dominant flow regime, the primary mechanism by which force is transferred is recoded as the flow progresses from initiation to deposition. We report on how different field measurable parameters, such as grain size distribution, slope of the bed, and flow thickness, influence the measured basal stress and we show how the basal stress changes as a function of the bed area over which it is measured.
H53C-1404
Debris flows as agents of landscape evolution in quartzite terrains of the central Appalachians, Virginia
Debris flows that impacted the central Blue Ridge Mountains of Virginia in 1995 provided unique insight into the landscape development of quartzite terrains. A pervasive pattern of bedrock joints and bedding planes forms orthogonal fracture systems that control denudation by debris slides and debris flows. Fracture recharge may serve to substantially increase rock pore pressure during high intensity rainfall, especially when coupled with antecedent moisture. The high-angle wedge-failure observed at debris-flow initiation sites possesses a similar geometry to that of the regional, mountainous landforms. The ubiquity of this geometry at all scales (i.e., outcrop, hillslope, and mountain) indicates that debris-flows are an important process in long term landscape evolution. Furthermore, the large volumes of sediment removed from relatively small basins indicate that debris flows are significant agents of denudation and sediment transport in the study region. Debris flows are episodic in this region as indicated by radiocarbon dating of a buried soil horizon, quantification of soil profile development, and stratigraphic relationships among debris-flow deposits. We estimate an approximate debris flow recurrence interval of 2000 to 4000 years for major regional events in this terrain during the Holocene.
H53C-1405
Potential Triggering Mechanisms for the 2006-2007 Half Dome Rockfalls, Yosemite National Park, California
Yosemite Valley is one of the most active areas of rockfall in the world, allowing for detailed examination of individual rockfall events. A rockfall database going back to 1857 reveals that more than half of all documented rockfalls were not associated with a recognizable triggering mechanism. Between July 2006 and June 2007, a series of at least eight rockfalls occurred from a single release point on the Northwest Face of Half Dome in eastern Yosemite Valley. The largest of these rockfalls occurred at 18:46 on July 27th, 2007, and had an approximate volume of 735 m3. Interestingly, all of the rockfalls occurred during the summer (June-August), with no apparent activity at the release point during the winter and spring, typically considered peak seasons for rockfall. In addition to mapping rockfall volumes and the distribution of rock debris, I investigated the geologic and hydrologic factors contributing to failure, including bedrock lithology, degree of weathering, joint density and orientation, and release point geometry. I also analyzed a number of potential rockfall triggering mechanisms, including earthquakes, precipitation, freeze-thaw, and thermal stresses. Although a number of factors contributed to weakening of the rock mass, no specific triggering mechanism(s) can be confidently linked to the rockfalls. Rather, the rockfalls likely resulted from progressive strain weakening of an overhanging arch, with initial small rockfalls destabilizing the rock mass to the point that a large failure occurred. The supposition that summertime rockfalls with unrecognized triggers are unusual has been used to support claims that rockfalls below Glacier Point were caused by wastewater discharges, but the 2006-2007 Half Dome rockfalls, which occurred in a wilderness setting, demonstrate that subtle, even unrecognizable, natural processes trigger summertime rockfalls in Yosemite Valley.
H53C-1406
Geomorphic Processes Along the Coastal Cliff, Central Coastline – Israel
A study aimed at finding the nature, rate and causes of the morphological changes in the coastal cliff in central part of Israeli coastline, was carried out between 1991 and 1996, along the coast of Michmoret – Giv'at Olga, Israel. The stratigraphical sequence of this cliff, which is of Pleistocene age, consists of alternating layers of Kurkar (a carbonate cemented quartz eolianite) and Hamra (a red brown sandy loam). The morphological changes occurring in the coastal cliff were documented, prior, during and immediately after storms, by field observation, redundant surveying and photography. The retreat of the coastal cliff is composed of several steps. As a rule the moisture level in the Hamra layers is higher than that of the Kurkar layers because of the higher content of silt and clay in the Hamra. During the rainy season the increase of moisture in the Hamra increases the relative weight and lowers its friction. This in turn intensifies the process of rill development in the Hamra layers and slides of material from this layer occur. As a result the overlying Kurkar remains without support, until the overhanging Kurkar breaks down and falls to the bottom of the cliff. The debris, which accumulates at the foot of the cliff, forms a moderate slope and a stable cliff profile, which prevents continuation of the erosion process. This does not last for long, because sea waves remove the debris from the foot of the cliff leaving it with a precipitous slope and vulnerable to further erosion. The magnitude of the landslide or rock fall event determines, to a certain degree, the frequency of the event. The larger the event, the less likely another landslide or rock fall will occur at the same point in the near future. Losses of soil and rill development at the cliff slopes are a result of runoff action. The wind action has an effect on the erosion of the coastal cliff by blowing the loose sand from Kurkar and Hamra layers. The main factors influencing the changes (retreat or forward movement) of the base of the cliff are: the rock strength at the base of the cliff, and the ability of the waves to reach the base of the cliff and erode or sweep away material. Sea waves break upon the base of the cliff only a few days during the year. The cliff base is protected by the Kurkar Beit- Yanai rocks, which fell from the top of the cliff. During the winter the beach profile is relatively lower and flatter than in the summer, with a difference of more then 120 cm. Therefore, it is in the winter that the sea waves have a greater ability to erode and sweep away material from the base of the cliff. The rock, debris fall and slides occur during the winter month, as a result of rainfall. It was found that when a rainstorm is smaller than 10 mm the probability of a rock or debris fall event are nil. The probability of rock, debris fall and/or slide increases with the amount and intensity of the rain. During a rainstorm of over 90 mm, the probability of a rock, debris fall and/or slide event is 100 percent. In the study area the main cause of the coastal cliff retreat is the rain. Sea waves have a secondary role in the cliff retreat, mainly that of sediment removal from the cliff base. Without the sediment removal from the cliff base, the cliff would reach a stable slope and retreat process would slow down or eventually stop. The average retreat of the coastal cliff in the study area during the five-year study was 24 cm/year.
H53C-1407
Evaluation of Model Coupling Frameworks for Use by the Community Surface Dynamics Modeling System (CSDMS)
The Community Surface Dynamics Modeling System (CSDMS) is a recently NSF-funded project that represents an effort to bring together a diverse community of surface dynamics modelers and model users. Key goals of the CSDMS project are to (1) promote open-source code sharing and re-use, (2) to develop a review process for code contributions, (3) promote recognition of contributors, (4) develop a "library" of low-level software tools and higher-level models that can be linked as easily as possible into new applications and (5) provide resources to simplify the efforts of surface dynamics modelers. The architectural framework of CSDMS is being designed to allow code contributions to be in any of several different programming languages (language independence), to support a migration towards parallel computation and to support multiple operating systems (platform independence). In addition, the architecture should permit structured, unstructured and adaptive grids. A variety of different "coupling frameworks" are currently in use or under development in support of similar projects in other communities. One of these, ESMF (Earth System Modeling Framework), is primarily centered on Fortran90, structured grids and Unix-based platforms. ESMF has significant buy-in from the climate modeling community in the U.S.; a closely-related framework called OASIS4 has been adopted by many climate modelers in Europe. OpenMI has emerged from the hydrologic community in Europe and is likely to be adopted for the NSF-funded CUAHSI project. OpenMI is primarily centered on the Windows platform and a programming language called "C-sharp" and is not oriented toward parallel computing. A third, DOE-funded framework called CCA (Common Component Architecture) achieves language interoperability using a tool called Babel. It fully supports parallel computation and virtually any operating system. CCA has also been shown to be interoperable with ESMF and MCT (Model Coupling Toolkit) and would appear to be interoperable with OpenMI if C-sharp language support were to be added to Babel. For all of these reasons CCA is very attractive as a base framework for the CSDMS project. This poster will present a comparison of several different coupling frameworks and will provide a demonstration of how component-based software can be developed within the CCA framework. http://csdms.colorado.edu
H53C-1408
Testing a new method for measuring micro-morphological change using embedded magnets in a travertine depositional environment, Fossil Creek, Arizona
Measuring micro-morphological change of bed topography in a fluvial setting is often done by installing erosion pins or drilling holes that serve as fixed reference points. A shortcoming to these methods is that protruding pins and open holes can influence the local processes that alter the channel bed. Here we report on the development of a new method for documenting micro-topographic change, using magnets embedded within actively forming travertine along Fossil Creek, Arizona. These natural travertine structures occur as channel-spanning dams that create a step-pool morphology that provides important aquatic habitat. Biotic processes in turn, such as microbial and algal growth and the trapping of floating leaves and branches, can catalyze travertine deposition and vertical growth in travertine dams. The goal of this overarching study is to document travertine growth rates, downstream of a recently decommissioned diversion dam, to determine the relative influence of various biotic and abiotic processes. The method consists of gluing individual magnets (1cm in diameter and thickness; 1gauss magnetic intensity) onto the end of 10cm lengths of PVC pipe with the magnetic poles parallel to the pipe. We then drill a vertical hole in either bedrock or travertine, place the magnet-pipe assembly into the hole, and back-fill the hole with crushed travertine, to approximate the pre-existing surface as best as possible. Prior to installing the magnets we measure the background magnetic field, using a Schonstedt GA-72CD magnetic locator, to ensure there are no nearby magnetic anomalies. Because the strength of the magnetic field decreases geometrically with distance, we created a calibration curve specific to these magnets, which is valid over a range of up to 40cm. We use repeat measurements of magnetic field intensity to document vertical travertine growth over time. We use repeat surveys with a total station as an independent check on the change in dam elevations, at and adjacent to the magnet locations. Although the basic method is simple, a number of challenges have arisen. There are differences in magnetic permeability of travertine, bedrock and air or water, which we are addressing with calibration curves for specific materials and dam geometries. Although we commonly installed magnets at local topographic high points along dam crests, non-uniform travertine growth can shift the location of dam crests, complicating efforts to re-occupy the exact magnet location for later measurements. We address this problem by searching for the highest intensity location in a horizontal plane above the dam crest, and by careful photo-documentation of the evolving travertine structures. So far, the method provides individual measurements of micro-topographic vertical position with a precision of 1 to 3mm, while elevation differences due to travertine growth are precise to within 1 to 2cm, depending on the distance to the embedded magnet and the material composition of the new increment of dam growth.