Longitudinal training walls: Optimization of river width subdivision

Recently, engineers propose longitudinal  training walls to replace traditional transverse  groynes. This new intervention is expected to  maintain a navigation route under low flow  conditions while not hampering flow  conveyance of the river channel. Navigation  occurs mainly in low-land river channels where  the formation of alternate bars constitutes a  problem which requires mitigation measures  like dredging. Le et al. (2015, 2016) found that  the starting point of the longitudinal training  wall with respect to a steady bar plays an  important role on the stability of the bifurcating  parallel channels. Starting at a location near  the upstream part of the bar leads to side  channel silt up. On the contrary, starting at a  location near the downstream part of the bar  leads to side channel erosion. The most  interesting result was that when the  longitudinal training wall starts near the bar top,  both channels remain open for a long time.  However, these results were obtained only for  a specific width ratio, ratio between the width of  the side and the width of the original channel,  B1/B0 = 1/6, under a constant discharge. In  practice, the width ratio may vary to obtain  specific achievements. Wang et al. (1995)  showed that the width ratio plays an important  role on the stability of bifurcating channels. So,  how the system behaves for different width  ratios under variable discharge remains  unclear and needs further investigation.