The conveyance capacity of compound river channels with urban scale flood plain roughness
Since Sellin (1964) reported the difficulties involved with discharge estimation in channels with floodplains, a significant amount of research has been undertaken into the hydraulic behaviour of compound channels. There is however a lack of empirical data relating to cases where urban development has taken place on the floodplain, given the large amount of urbanisation that has taken place on the banks of many rivers, it is obvious that this is an area of research that needs attention.
This investigation addressed this need by investigating a compound channel with roughness simulating an idealised urban area on the floodplain. The aim was to quantify the effects of changes in roughness configuration on discharge and flow resistance.
The research programme included 202 tests utilising a compound cross section and 129 tests in a channel of rectangular geometry. Symmetrical and asymmetrical arrangements of roughness were investigated in the compound channel, with simple channel testing providing a data set against which compound channel results could be compared. The longitudinal bed slope was set at 1.85x 10-3 for all compound and simple channel test cases.
The research highlights the effects of large scale vorticity, previously unrecognised in floodplain research. This vorticity acts with several other factors and has the effect that, when the roughness element length to streamwise spacing length ratio was unity, the roughness coefficient of the total channel reduced as rows of roughness were added behind an initial row of roughness placed at the main channel edge. This was found to be caused by increases in the main channel and smooth floodplains capacity. A consideration of this vorticity together with other reasons for observed behaviour has been presented. An extensive sensitivity analysis has been carried out considering the effect varying the longitudinal and transverse roughness element spacing has on these phenomena.