Measured channel resistance at mountain rivers during flood

Yuko ASANO, Taro UCHIDA and Yuki NISHIGUCHI

Flow resistance coefficients, such as Manning's roughness coefficient n have been useful, however, property of flow resistance at steep mountain streams are not well understood because of lack of data. We have found that resistance dramatically decrease with increase of discharge but once the water surface was above most of the boulders in the bed, the coefficient then changed little with increasing water depth. If this can be a common phenomenon, prediction of flood flows can be possible using resistance equation that assumed resistance does not change with water depth even at mountain channel. We collected existing data from literature that documenting changes in flow resistance during flood and (1) examined the relationships between relative water depth and channel resistance for each channel, (2) extracted the smallest resistance from each channel and demonstrated the measured ranges of resistance according to channel morphology, and (3) investigated the effect of slope, catchment area and grain size on flood flow resistance. We collected data from 76 channels with bed slope over 1/100. They were classified as cascade, step]pool, plane]bed and pool]riffle morphology. Manning's n decreased with increasing relative water depth, and for most of channels, resistance changed little with increasing water depth on higher water levels. Measured minimum Manning's n, that should represent the resistance during flood, ranged from 0.03 to 0.35. Minimum Manning's n increased with slope and grain size for plane]bed and pool]riffle channels, while no such relationships can be found for cascade and step]pool channels. These results suggested that for the prediction of the large flood flows, constant resistant values can be used also for mountain channels. The prediction accuracy of channel resistance during flood can be improved using slope and grain size distribution for plane]bed and pool]riffle channels but not for cascade and step]pool channels. For better prediction of resistance for cascade and step]pool channels, we may need more information that should contribute to channel resistance, such as bedform.

Key wordsFCascade, channel morphology, Manning's roughness coefficient, small stream, step]pool