| dc.description.abstract | Urbanization has a major impact on flow and erosion in open channels. By replacing rural land surfaces with impervious surfaces, urbanization causes a decrease in infiltration and an increase in surface runoff and faster routing to the channels. For two identical rain events on both rural and urban catchments, more water enters urban channels in a shorter period of time, which means that flow hydrographs [graphs that show changes of flow discharges over time] in urban channels are characterized by shorter times to peak, higher peak flows, and rapid changes in flow depth and discharge. However, while such rapid changes are known, they are not typically considered for the purposes of floodplain delineation and risk, which usually use steady flow assumptions. There are limited studies that focus on characterizing unsteady flows in urban catchments and the degree to which unsteadiness affects the channel hydraulics, i.e. the balance of forces on the water, bed and banks of a river that determine the flooding, sediment transport, and erosion during a flood. Numerous damaging floods occurring annually around the world shows that floodplain maps have not been delineated accurately. Urban floods also cause severe damage to nearby infrastructure in rivers and floodplains due to massive fluxes of sediment transport occurring under unsteady urban flows. Although it is known that sediment transport rate is high in urban channels and that the hydrographs are flashy, there is little research on sediment transport in urban channels. Due to the limitations and complexity of bedload transport, most studies investigate it under steady flows. Therefore, sediment transport rate under such unsteady flows requires a comprehensive further research.
The first objective of this thesis is to investigate the level of unsteadiness of flows in urban catchments versus rural catchments. For this purpose, we selected two urban catchments one with and one without Storm Water Management (SWM) facilities. We also selected one rural catchment with similar shape, watershed slope, and drainage area. All three watersheds are located near the City of Toronto not far from each other. Pressure gauges were installed to monitor flow depths with frequency of 2 minutes during summer months. This data was used to route flows downstream using the Saint Venant equations. By looking at both hydrologic and hydraulic parameters, it can be concluded that adding SWM facilities to urban creeks dampens hydrologic parameters, but not hydraulic parameters. In addition, it can be seen that the flows observed in the urban catchment have higher level of unsteadiness compared to that in the rural catchment.
The second objective of the thesis is to assess the accuracy of using a system of two gauges that measure the water surface elevation to predict, or ‘route’, the water discharge through an urban creek characterized by floods with high flood magnitudes and rapid rising limbs. We installed pressure gauges with 15-second frequency in a watershed with a high percentage of impervious land cover and used four different methods of flow routing to calculate discharge. The results were compared with direct discharge measurements to evaluate all the routing methods. The results of this chapter showed that lower frequency data in the urban channels cannot represent the high degree of unsteadiness of flows and that in comparison with a standard approach that uses one gauge measuring every 15 min, a two-gauge system measuring water surface elevation every 2-5 minutes would reduce the associated error by 20-30%. One-gauge monitoring systems remain the standard for developing monitoring stations for the purpose of establishing flow rating curves in most channels, but the results show that a two-gauge system would reduce error for the studied rural and urban rivers.
The final objective of the thesis is to investigate sediment transport under the types of unsteady flows that are observed in urban areas. Physical experiments were run in a scale model of an urban creek investigated in Chapter 2 and Chapter 3 of the current thesis. The results from experiments with a set of triangular hydrographs were compared to assess the impact of the rising limb duration on the morphology and sediment transport in the channel. The hydrographs all had the same duration and peak magnitude. Mixed sediment sizes were used for this project based on the actual sediments observed in the creek. The bed was covered evenly over the entire flume with a mixture of sand and gravel and no additional sediment inflows were added during the experiments. Three large sediment size ranges were each colored with a different fluorescent color. The movement of sediments was captured by a high-resolution camera under black lights, and tracked using an available object identification and motion detection algorithm. We show that the shape of the hydrograph is only of secondary importance as there is more variability between experiments with the same hydrograph than those with different hydrographs. This result indicates that the complexity of the bed sediment interactions was likely controlling export from the flume and that there is a random or chaotic component of the sediment transport. Analysis by particle size showed that there were some clear trends in sediment export and that the coarse sediment tended to be over represented in sediment output during the high peaks. Where the rising limb was rapid, the coarse sediment was even more over-represented in the output, which makes it surprising that rising limb rate was not the main driver in the difference of overall sediment export. These results support the need for additional research on the interaction between the flow, transport and morphology. In particular the tracking techniques should better allow the spatial details and differences between the experiments to be extracted.
Overall, the thesis concludes that unsteady flows are observed both in urban and rural channels, but are more frequent in urban channels. There will be lower degree of errors for discharge estimation in unsteady floods if the two-gauge routing method presented in this thesis is used. Sediment transport under such unsteady flows require further investigation with the tracking method presented in this thesis. Therefore, additional laboratory and field research is required for the purpose of better understanding the characteristics and effects of such unsteady flows on the surrounding environment. | en |
