河川における側岸侵食

Abstract

Observations of side bank erosion in rivers are very important, as well as experimental andtheoretical investigations, to estimate the safety of river banks and the stability of river channels.Bank erosion data in low water courses in Teshio River and Uji River are mentioned as examples.Annual changes of the cross sectional shape at 43 km site in Uji River are analysed by usingdischarge data at Yodo Water Gauging Station, 4 km downstream.Bank erosion processes have a large spacial difference and some intermittency. In channelswhere bank erosion continues, erosion becomes rapid when the mean shear stress comes up near20 N/m2 while erosion stops when it becomes less than 10 N/m2. The intermittency indicatesthat banks containing silt and clay have also the erosion cycle as same as in experiments: scour ofthe bank slope-collapse of the upper part-fluvial remove of the deposit part. Bank erosion isaccelerated by local bed scour near the bank slope.An observation of bank erosion process in a low water course has been carried out since June, 1980 near 43 km site in Uji River. Outlines of the observation are shown and the results arediscussed mainly on the shift of bank edge line and its receding rate, being compared not onlywith flood discharge and duration but also with hydraulic quantities, which are calculated fromwater stage and discharge data of Yodo and Mukaijima Water Gauging Staion.The receding rates correspond to the flood discharge and the total duration as a whole, butthe relation between them is not clear. The difference in the erosion rates seems not so large asthat in the discharge. One of this reason is that the shear stress does-not increase monotonouselyaccording to the discharge because of the channel conditions, such as backwater from the junction.Actually, at the peak of the flood in August, 1982, the shear stress became smaller than 7 N/m2 inspite of the discharge of more than 1000 m3/sec, because of the decrease of the water surface slopedue to the backwater. The bank erosion rate was, therefore, found to correspond well to thevalues and the duration of shear stress by the hydraulic analyses of the floods in 1980 and 1982.