Hydrology [H]

 CC:Hall E  Sunday  0800h

Models and Measurement of Sediment Transport Posters

Presiding:  M Van De Wiel, University of Western Ontario; L Gordon, University at Buffalo; D De Boer, University of Saskatchewan; J Tunnicliffe, Carleton University


The Fate of Organic Carbon Released by Permafrost Decay, Eastern Coast of Hudson Bay.

* Jolivel, M (maxime.jolivel.1@ulaval.ca), Centre d'Etudes Nordiques, Pavillon Abitibi-Price, local 1229. 2405, rue de la Terrasse Universite Laval, Quebec, Qc G1V 0A6, Canada
Allard, M (michel.allard@cen.ulaval.ca), Centre d'Etudes Nordiques, Pavillon Abitibi-Price, local 1229. 2405, rue de la Terrasse Universite Laval, Quebec, Qc G1V 0A6, Canada

Recent evaluations indicate that large amounts of organic carbon can be released in fluvial and costal systems because of permafrost degradation, with impacts on ecosystems. In order to obtain quantitative data on those transfers, we have installed instrumentation and have made first measurements in an intensive permafrost degradation area. The study area is located on the eastern coast of Hudson Bay, in the region of the Sheldrake river (drainage basin, river mouth and offshore area), near Umiujaq, in the discontinuous permafrost zone. Permafrost mounds (palsas, lithalsas) and plateaus are the most abundant permafrost landforms. This area contains one of the largest concentrations of frost heave landforms in the world. They developed principally in east-west oriented valleys in postglacial marine silts from the Tyrrell Sea which inundated low areas around Hudson Bay, following the receding ice front eastward and inland, about 8000 years BP. Palsas are covered by peat. Organic matter and clay released by thermokarst are transferred to the sea through the river system as suspended sediments, suspended organic matter and dissolved organic carbon. We postulate that continuing warming will further accelerate permafrost erosion, favour thermokarst and have an impact on carbon transfers. The adopted methodology should permit to quantify the release of clay and carbon through fluvial transport and deposition in coastal marine depocenters. Two leveloggers and two Optical Backscatter Sensors (OBS) have been installed during summer 2008, about 2km upstream from the Sheldrake river mouth in order to estimate transportation. Moreover, bathymetric surveys (eco-sounder coupled with sidescan sonar) have been made in a 20km2 area offshore the Sheldrake river mouth. We have located deeper basins (88m deep) in submerged valleys, which are likely efficient sediments traps for recent river inputs. Some sediment cores will be extracted in March 2009 from the ice pack in order to correlate recent marine sedimentation with erosion in the drainage basin and permafrost degradation. Grain-size studies, C-14 dating, use of 137Cs, 210Pb, C/N ratios and δ13C will allow to ascertain carbon sources and calculate sedimentation rates.


Integrating Geomorphological and Ecological Processes in Numerical Landscape Evolution Models

* Van De Wiel, M J (mvandew3@uwo.ca), University of Western Ontario, 1151 Richmond Street, London, ON N6A 5C2, Canada

This paper outlines the framework for a new research programme which aims to improve understanding of interactions and feedbacks between ecological and geomorphological processes in landscape evolution. Landscapes are complex dynamic systems where geomorphological processes are influenced by the landscape's vegetation cover and vice versa. Given these mutual relations it can be expected, and indeed it has been shown, that geomorphological and ecological processes interact and co-determine the evolution of the landscape. However, these interactions are currently poorly understood and usually neglected in numerical landscape evolution models. Hence, a new landscape evolution modelling approach needs to be developed, integrating ecological processes and geomorphological processes, which permits to study the impacts of these processes and their interactions on spatial patterns of erosion and deposition, as well as sediment yield, sediment storage and overall landscape evolution. Here, a distinction is made between short-term (annual to decadal) and long-term (centennial to millennial) landscape evolution, and two corresponding modelling approaches can be considered: a short-term model applied to small spatial scales (e.g. smaller subcatchments); and a long-term model applied to larger spatial scales (e.g. larger sub-catchments or entire drainage basins). The approach outlined here attempts, where possible, to integrate empirical research for informing model development, as well as for validating model predictions in real-world scenarios.


A New Monitoring Method of Individual Particles During Bed Load Transport in a Gravel Bed River.

Tremblay, M (michele.tremblay.3@umontreal.ca), Chaire de recherche du Canada en dynamique fluviale, Département de géographie Université de Montréal C.P.6128, Succ. Centre-Ville, Montréal, Qc H3C 3J7, Canada
Marquis, G (ge.marquis@gmail.com), Chaire de recherche du Canada en dynamique fluviale, Département de géographie Université de Montréal C.P.6128, Succ. Centre-Ville, Montréal, Qc H3C 3J7, Canada
Roy, A (andre.roy@umontreal.ca), Chaire de recherche du Canada en dynamique fluviale, Département de géographie Université de Montréal C.P.6128, Succ. Centre-Ville, Montréal, Qc H3C 3J7, Canada
* Lamarre, H (helene.lamarre@umontreal.ca), Chaire de recherche du Canada en dynamique fluviale, Département de géographie Université de Montréal C.P.6128, Succ. Centre-Ville, Montréal, Qc H3C 3J7, Canada

Many particle tracers (passive or active) have been developed to study gravel movement in rivers. It remains difficult however to document resting and moving periods and to know how particles travel from one sedimentation site to another. We have developed a new tracking method using the Hobo Pendant G acceleration Data Logger, to quantitatively describe the motion of individual particles from the initiation of movement, through the displacement and to the rest, in a natural gravel river. The Hobo measures the acceleration in three dimensions at a chosen frequency. Hobo Pendant G Acceleration data logger were inserted into 11 artificial rocks and seeded in Ruisseau Béard, a small gravel river in the Yamaska drainage basin (Québec). The hydraulics, particle sizes and bed characteristics of this site are well known. Controlled tests have been performed before the field experiment to understand the response of the instrument. The results allow us to develop an algorithm which classifies the signal into periods of rest and motion. The algorithm can also differentiate the type of motion: vibration, rolling and sliding of the particles. The data allow us to describe the time of movement, the path length and the velocity of the particles. The comparison of the movement and rest periods to the hydraulic conditions (discharge, shear stress, stream power) established the movement threshold and response times. Relations with bed roughness and morphology were also established. Finally, the development of a 2-dimension model helps visualizing the angular variation motion and a 3D model allows the reconstitution of the particle trajectories on the bed. This method offers great potential to track individual particles and to study bedload transport in rivers. This first attempt needs to be further improved especially to retire the degree of precision of the movement detection. The method should also be tested with frequencies higher than one minute, with more particles of different shapes and sizes.


Erosion and Surface Structures in the Initial Phase of Ecosystem Development Investigated in an Artificial Catchment

Raab, T (raab@tu-cottbus.de), Brandenburg University of Technology Cottbus, Chair of Soil Protection and recultivation, Konrad-Wachsmann-Allee 6, Cottbus, D-03046, Germany
* Gerwin, W (werner.gerwin@tu-cottbus.de), Brandenburg University of Technology Cottbus, Research Center Landscape Devlopment and Mining Landscapes, Konrad-Wachsmann-Allee 6, Cottbus, D-03046, Germany
Dimitrov, M (dimitma@tu-cottbus.de), Brandenburg University of Technology Cottbus, Research Center Landscape Devlopment and Mining Landscapes, Konrad-Wachsmann-Allee 6, Cottbus, D-03046, Germany

In the very first stage of the ecosystem development hydrological processes forming surface structures are mainly controlled by runoff patterns and by physical properties of the substrate. Based on that, it can be hypothesized that the initially formed structures are responsible for the future development of the ecosystem and define later structures. However, initial structures are very dynamic, and few alterations of surface properties may initiate the development of completely new patches and patterns which again control surface processes like erosion and sedimentation. For example, a first physical soil crust is formed very quickly and exhibits clearly different properties compared with the original initial surface. This influences infiltration and surface runoff. In contrast, substrate that has been deposited changes the soil physical properties into another direction. In these sedimentation areas small patches with higher infiltration rates and eventually better water storage capacities may be created. This may result into the formation of initial vegetation patches and patterns which in turn influence the further quality and quantity as well as the location of soil surface processes. This paper presents a recently launched research project using an artificially created water catchment of 6 ha in size. This site called Chicken Creek (Huehnerwasser) was established in 2005 in Lusatia (Germany) and is the central research site of a German-Swiss Collaborative Research Centre. The catchment was designed as a landscape laboratory representing an initial ecosystem starting its development at point zero. In the first three years of this development already several surface structures and their interactions could be detected. As it was expected, the initially not vegetated area showed massive substrate dislocation by erosion and the formation of numerous gullies. Monitoring of erosion processes as well as the analysis of the interactions and the complex feedback of these different structures will be an important issue for the Collaborative Research Centre. This paper presents first results of the monitoring programme and concepts of future measurements.



Quantification and Modelling of Fugitive Dust Emissions From Nickel Slag

* Sanderson, R S (robertsanderson@trentu.ca), Trent University, Department of Geography, 1600 West Bank Drive, Peterborough, ON K9J 7B8, Canada
McKenna Neuman, C (cmckneuman@trentu.ca), Trent University, Department of Geography, 1600 West Bank Drive, Peterborough, ON K9J 7B8, Canada

Mining and smelting operations in Northern Ontario, and indeed worldwide, introduce a number of unique sources of fugitive dust and other aerosol pollutants into the surrounding environment from smokestacks, tailings, and slag dumps exposed to wind erosion. Fugitive dust represents a potential health hazard, and as such, mining companies are required to maintain inventories of dust emissions associated with their operations. The purpose of this study was to fully characterize the wind-induced fugitive dust emission rates of nickel slag collected from a slag dump at a smelting facility in Northern Ontario, as dependent on wind speed, surface roughness, duration of weathering, effects of mechanical disturbance, and exposure to rain. PM10 flux rates were measured through combined field monitoring and wind tunnel simulation. In both settings, airborne dust concentrations downwind of the source were measured using four vertically distributed DustTrak aerosol monitors. Wind speed was measured in the wind tunnel using a micro-pitot tube mounted on a programmable traversing slide, and in the field, using five vertically distributed cup anemometers mounted on a mast. The profiles of PM10 and wind speed were used to compute the vertical emission rate (Fv) using a finite difference method. The PM10 emission rates simulated in the laboratory were found to directly overlap those measured on site at the smelting facility over a range of wind speeds, suggesting that Fv values measured in wind tunnel simulations can be used in dispersion modelling with a reasonable degree of confidence. Although showing a strong positive correlation with wind speed, PM10 emissions from nickel slag were found to demonstrate an exponential, temporal decay immediately following any form of mechanical disturbance that resulted in exposure of the silt fraction of the material. Winnowing of this fraction left behind an armoured surface of coarse, non-erodible clasts. It was further determined that percolation of rainwater can transport the silt fraction several centimetres down into the deposit, so that for an undetermined time afterward, Fv is reduced by approximately one order of magnitude. In the long term, winnowing and eluviation processes acting on an undisturbed area may produce an essentially non-emissive surface.


Stochastic Modeling of Non-equilibrium Bedload Transport

* Kuai, Z (kzheng@buffalo.edu), University at Buffalo, 212 Ketter Hall, Buffalo, NY 14260,
Tsai, C W (ctsai4@eng.buffalo.edu), University at Buffalo, 212 Ketter Hall, Buffalo, NY 14260,

Traditional stochastic bed load models aimed to solve for the equilibrium bedload transport rate by matching the rate of bed erosion with the rate of deposition. Bedload transport can be in nonequilibrium even under the steady flow condition, as the quantity of moving particles in the bedload layer may vary. In a nonequilibrium condition, the interchange of sediment particles occurs not only between the bedload layer and the bed surface, but also across the interface between bedload and suspended load. The proposed approach attempts to add a new bedload-suspended load interchange layer to a stochastic bedlod transport model based on the Markov chain. The bedload transport rate is the product of the total particle volume in saltation and the average saltating velocity. We can quantify the number of saltating particles by modeling the occupancy probabilities vector of particles staying in three states (i.e., bed surface, bedload layer, and the interchange layer between the bedload and the suspended load.). The new stochastic bedload relation is validated against existing bedload model. The sudden change of flow and/or sediment condition leads to changes in the transition probabilities. The influence of sudden changes in flow-sediment properties on the bedload transport rate is investigated in this preliminary study. It is found that the neglecting the exchange process between the bedload layer and the suspended layer may lead to non-negligible errors in bedload calculation when the flow and/or sediment conditions change.


Sediment Particle Movement in Open Channel Flows Using a Stochastic Jump Diffusion Particle Tracking Model

* Oh, J (joh8@buffalo.edu), State University of New York at Buffalo, 207 Jarvis Hall, Buffalo, NY 14260, United States
Tsai, C W (ctsai4@eng.buffalo.edu), State University of New York at Buffalo, 233 Jarvis Hall, Buffalo, NY 14260, United States

Movement of sediment particles in flows is generally led by the flows. Particles being transported in open chanel flows can vary from suspended load, wash load to bed load. Deposition occurs when the velocity of movement of the transporting medium becomes insufficient to hold the particles. On the other hand, erosion or resuspension occurs when the flow velocity becomes sufficiently large to move the particles. In particular, extreme flow events increase the unsteadiness of flows so that local erosion and deposition are severely intensified and the effect on geomorphologic changes is enhanced. Despite the enormous impacts of extreme flow events on sediment transport, movement of sediment particles in response to extreme flow events in open channel flows has not been fully understood. Herein, in this study, we propose a model with a particle-based Lagrangian approach so that we can trace and predict the movement of sediment particles. Hence, the model is capable of realizing the estimated pathway of particles with or without extreme flow events and quantifying the influence on channel beds, i.e., bed change. Futhermore, the proposed model also adopts a stochastic approach. The stochastic jump diffusion processes in the proposed model can consider inherent randomness of movement of sediment particles. The stochastic differential equation (SDE) accounts for paticle movement in flows with three parts: mean drift movement, random tubulent movement and jumps due to extreme flow events. A stochastic diffusion process is applied to describe random motion of sediment particles due to flow turbulence. Abrupt movement of sediment movement due to extreme flow events is modeled by a stochastic jump process. The Weiner process represents the stochastic diffusion process; and the Poisson process characterizes the stochastic jump process. Moreover, in the proposed model, deposition and resuspension of sediment particles are determined by stochastic bed shear stress compared to the critical shear stress. The stochastic property of bed shear stress originates from the stochastic flow velocity. A few examples show the movement of sediment particles released at the center of the depth or initially stayed on the bed in open channel flows. As the result, the proposed stochastic jump diffusion particle tracking model (SJD-PTM) can model the movement of sediment particles in two-dimensional open channel with time. The most probable trajectory of sediment particles can be displayed by the ensemble mean of particle trajectories. Uncertainties of sediment particle trajectories can be described by the ensemble variance of particle trajectories. Besides, particle deposition / resuspension represented by the stochastic bed shear stress indicates that extreme flow events in open channel flows tend to let sediment particles resuspended. It is demonstrated that the estimated bed shear stress is sufficiently increased to exceed the critical shear stress due to increased flow velocity during extreme flow events. It is shown that particles are likely to be deposited on the channel bed in regular flow and resuspended by obtaining the velocity by extreme flow events. Accumulation of these results of a number of particle movements can reflect deposition and erosion on the river bed. The results demonstrate that the stochastic diffusion jump particle tracking method can be used for identifying the original position of sediment particles or inferring whether or not extreme flow events occur. Therefore, the proposed stochastic jump diffusion particle tracking model (SJD-PTM) can offer comprehensive prediction for sediment transport.