Experimental study on evolution of bed structures of natural mountain rivers

Huai-xiang LIU*, Zhao-yin WANG, Guo-an YU, Kang ZHANG

1. State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Nanjing Hydraulic Research Institute, Nanjing 210029, P. R. China

2. State Key Laboratory of Hydroscience and Engineering, Tsinghua University, Beijing 100084, P. R. China

Experimental study on evolution of bed structures of natural mountain rivers

Huai-xiang LIU*1, Zhao-yin WANG2, Guo-an YU2, Kang ZHANG2

1. State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Nanjing Hydraulic Research Institute, Nanjing 210029, P. R. China

2. State Key Laboratory of Hydroscience and Engineering, Tsinghua University, Beijing 100084, P. R. China

Bed structures in many mountain rivers provide additional resistance to the flow. A field experiment was conducted on debris flow deposits in the valley of the Jiangjiagou Ravine, a tributary of the Yangtze River in southwestern China, to study the evolution and distribution of bed structures and their relationship with environmental conditions. Water and sediment from the Jiangjiagou main stream were diverted into the experimental channel. Several hydrological schemes were adopted to scour the channel until equilibrium was reached. During this process the evolutions of bed structures and channel configuration were investigated. The results indicate that stronger bed structures mean greater stream power consumption, greater resistance, and greater slope in a certain section when rivers are in dynamic equilibrium. Thus, to some extent the longitudinal profiles of channels can be determined by the distribution of bed structures. In natural cases, the strength and evolution of bed structures are under the influence of environmental conditions such as discharge and bed-load transportation rate.That is, given the same conditions, the same bed structure distribution and longitudinal profile canbe predicted.

bed structures; channel configuration; channel incision; slope; resistance

1 Introduction

River bed configurations in natural streams are shaped by varying flows. At competent flows the least stable particles move into more stable positions to create structured bed forms. Commonly occurring bed structures include the step-pool system, ribbing structures, and boulder and cobble clusters (Wang and Lee 2008). The step-pool system (Fig. 1(a)) is a geomorphologic phenomenon with alternating steps and pools and a stair-like appearance (Chin 1999; Wang et al. 2009). It develops usually in small mountain streams with a several-meter wide channel. Cobbles and boulders generally form the steps, which alternate with finer sediments in pools to produce a repetitive staircase-like longitudinal profile in the stream channel. Cobbles and boulders overlap with each other and form ribs extending out from the banks. This structure exhibits high stability and enhances resistance against the flow,thus protecting the banks and channel bed from erosion. The Balan River (Fig. 1(b)) is a tributary of the Songhua River in northeastern China. The bed slope is about 0.5% to 1%, and ribbing structures have developed on the bed. The ribbing structures are composed of cobbles, gravels, and some boulders. The structures have made the channel incision much slower, and the channel bed is now quite stable. Boulder and cobble clusters are accumulations of sediments on either (or both) the lee or stoss side of an obstacle clast in rivers with poorly sorted sediments (Wittenberg 2002; Strom et al. 2006). They are the most prevalent type of small-scale bed structures in gravel-bed rivers and contribute to roughness properties and the enhanced bed stability. Fig. 1(c) shows the boulder clusters in the Baohe River, a tributary of the Hanjiang River. There are also other types of bed structures such as star-studded boulders and bank stones (Wang and Lee 2008).

Fig. 1 Common bed structures in mountain streams

The bed structures have been observed in a wide range of humid and arid environments (Chin 2002), and similar forms have even been observed in glacial streams. Thus they appear to be a fundamental element of steep fluvial systems. There have been many studies on regular bed structures. Several variables are usually used to characterize the morphology of bed structures, including the mean structure heightH, the mean structure length (distance between two individual structures)L(Fig. 2), channel slopeS, and so on. Researchers have used these variables to quantify the bed structures and study the relationship between these geometrical parameters and environmental conditions. Rosport (1997) reported that the lengthLof a regular step-pool system increases with the average discharge. Whittaker (1987) suggested that the lengthLof a regular step-pool system, or the distance between two steps or two pools, is inversely proportional to the average slope of the stream:

whereLis in meters. The decrease inLis rapid as the slope increases up to about 0.15. The influence of bed slope on the adjustment of step-pool morphology is further illustrated by a relationship between the average step steepness () and slope obtained by Abrahams et al. (1995) from field and laboratory data:

This relationship indicates that the average elevation loss due to steps is about one to twotimes that along a reach, which implies that about one third of the step height is the result of pool scour. The experimental results of Wang et al. (2004) show that Manning’s roughness coefficientnlinearly increases with the scale of bed structures.

Fig. 2 Bed structures and their parameters

Generally, the functions of bed structures are to enhance flow resistance, to consume flow energy, and to prevent streambed erosion (Abrahams et al. 1995; Yu et al. 2008; Kim et al. 2010). The bed structures can reduce the lift and drag forces acting on the particles on the lee side of the structures (Reid 1992; Yang et al. 2006). In streams in the Cascade Mountains of Washington State, the resistance caused by the step-pool systems accounts for more than 90%, while the grain resistance and channel form drag make up only less than 10% of the total (Curran and Wohl 2003). Sear (1996) found that clusters are among the configurations of a gravel river bed most resistant to entrainment. The shear stress required to entrain clustered sediments is higher than that required to entrain dispersed bed particles (Hassan and Church 2000).

The investigation results of natural rivers were mainly applicable locally due to restricted environmental conditions. The previous experiments were mainly conducted in the laboratory, and could not simulate the natural evolution process perfectly. A systematic study of bed structures is still required. This study designed a field experiment located in a natural river valley, in order to analyze the evolution and distribution of bed structures under different environmental conditions, as well as the underlying mechanism of fluvial morphology.

2 Jiangjiagou field experiment

The Jiangjiagou Ravine is a tributary of the Xiaojiang River (a tributary of the upper Yangtze River). With a drainage area of 48.6 km2, and a main channel length of 13.9 km, this ravine is known as a natural museum of debris flows. Its elevation ranges from 1 042 m to 3269m, resulting in a steep landform. The vegetation coverage is quite low but the precipitation is relatively high and is concentrated in rainy seasons, leading to frequent flash floods and debris flow events. Every year at least 6 × 106m3of sediment is discharged into the Yangtze River due to debris flows occurring in the Jiangjiagou Ravine. During this process, a large volume of sediment is also deposited in the Jiangjiagou Valley. Therefore, over manyyears, the valley’s bottom has been filled with debris flow deposits. These deposits are distributed as flat layers overlapping according to time orders. Deep layers consist of old remains brought by debris flows, while layers nearer to the surface are newer.

The debris flow deposits contain particle components with a wide grain size spectrum, from clay to boulders, which is not common for general land cover materials. Because sediment carried by debris flows is poorly sorted, it can provide bed structures with various sediment resources. This means that this is an ideal place for the field experiment examining bed structure evolution and distribution. In this experiment, the surface of the plane of debris flow deposits is assumed to be the original land surface without hydraulic erosion. Water and sediment are diverted onto the deposits for simulation of rivers so that free degradation, aggradation, and evolution of bed structures can be anticipated and observed.

The field experimental layout is shown in Fig. 3 and Fig. 4. On the surface of debris flow deposits artificial shallow channels were pre-made for the diversion of water and sediment from the main stream of the Jiangjiagou Ravine, and also for flow direction regulation. The artificial channels was then reconnected the Jiangjiagou Ravine downstream to form the exit of the experimental section. Since the elevation at the downstream confluence was relatively stable, an erosion base level existed for channel incision in the debris flow deposits.

Fig. 3 Field experiment in Jiangjiagou Valley

Fig. 4 Layout of field experiment (phase 1)

In the artificial channel a small cement Parshall flume-type flowmeter (Fig. 4) was made to estimate the discharge diverted. Fifteen cross-sections were set along the channel, named 1#, 2#, 3#, etc., from downstream to upstream, as shown in Fig. 4. An electronic theodolite was chosen to monitor elevation change in both longitudinal and cross-sectional profiles. Some other hydrological measurements were taken during this experiment, including point velocity (measured by a pitot tube), suspended load concentration (measured by taking water samples), bedload transportation rate (measured by a sedimentation tank), and grain size distribution of sediment (measured by the sieving method).

The basic process of this Jiangjiagou experiment is described in Table 1. The operation time was divided into three phases. For every phase, an artificial channel was pre-made at a different location, and different hydrological schemes were adopted. There were two basic hydrological schemes, the bedload scheme and non-bedload scheme, running alternatively. The bedload scheme means that the water and sediment diverted from the upper Jiangjiagou Ravine were poured into the experimental reach without any treatment, while the non-bedload scheme means that the diverted bedload was removed by a sedimentation tank dug adjacent to the Parshall flume’s upper end (Fig. 4). Every scheme started after the previous scheme had reached dynamic equilibrium. Every phase repeated the former phase first and then took one more run of the bedload or non-bedload scheme. The diversion process aimed at simulating natural hydrological series. The diverted discharge was controlled in such a process that high-water and low-water periods both existed (Fig. 5). Phase 1 and phase 2 were totally“natural” in the sense that they simulated free erosion and sedimentation without any human intervention. However, in phase 3 some artificial structures were placed in the channel in order to estimate humans’ capability of changing fluvial morphology.

Table 1 Basic process of Jiangjiagou experiment

Fig. 5 Diverted discharge process

As shown in Fig. 4, the erosion base level is lower than the surface plane of debris flow deposits by 1 m to 1.5 m (cross-sectionAB). Thus in every phase the experiment started with severe retrogressive erosion. The channel incised rapidly and this effect gradually spread upward. After a period of bed evolution, bed structures evolved in certain arrangements on the river bed and stabilized the channel. Fig. 6 shows that after the incision fine particles (with a grain sizeDless than 10 mm) had almost been scoured and removed completely. In Fig. 6,Wis the weight ratio of particles with a size less than the grain sizeD. Only coarse particles could remain on the channel bed and become components of bed structures. Although the scales of structures differed among phases and hydrological schemes, they could still be grouped mainly into three types according to their configuration: step-pool, ribs, and clusters.

Fig. 6 Grain size distribution of deposits, bedload, and bed structures

3 Results and discussion

3.1 Bedload scheme

The bedload scheme ran at least once in every phase (Table 1). From phase 1 to phase 3, the bedload scheme was always chosen as the first one to scour the artificial channel. In phase 3 there was even a second run of the bedload scheme as the final step after a non-bedload scheme. Table 2 gives the shape parameters of all initial artificial channels. It was obvious that the channels were very different in many aspects such as plane view and cross-section at the beginning. For instance, in phase 1 the pre-made channel was meandering with several bends (Fig. 4), while in phases 2 and 3 the channels were almost straight. Different cross-sections (U-form/V-form, and with different width/depth ratios) were chosen for the three phases.

Table 2 Initial geometric properties of artificial channels

The initial channels were deliberately made very different in every aspect. However, as shown in Fig. 7, their longitudinal profiles in equilibrium status almost superimpose on each other under the same hydrological conditions. In Fig. 7, phase 3(1) and phase 3(2) represent the first and second bedload schemes in phase 3, respectively. The differences between their slopes were below 10%. The statistics listed in Table 3 describe the bed structure evolution in experimental channels in dynamic equilibrium status of the bedload scheme, as well as their relationship with slope. Within the same reach the evolved bed structures were also quite similar among different phases in both type and size. That is to say, the evolution and distribution of structures are in close relation to the hydrological conditions.

Fig. 7 Equilibrium profile in bedload scheme

Table 3 Bed structures and slopes in bedload scheme

As for river patterns, all channels became wandering rivers with no fixed main stream (Fig. 8). The most prominent feature was that the streams changed so frequently that the river bed was widened to a shallow U-form one and the width reached nearly 10 m (Fig. 8). According to the field records, the channel thalweg kept migrating to other branches. Thus, the banks on both sides were eroded alternatively and vast sand bars emerged within the range of migration. In Fig. 9 this kind of channel widening can be easily seen.

An increasing trend of bed structure height and length from upstream to downstream can be seen in Table 3. The reason is that the grain size remaining on the downstream river bed after scouring of the bedload scheme was larger than that upstream on average (Fig. 6). Downstream the weight ratio of the grain size (D) smaller than 100 mm was greatly reduced, indicating that this group of sediments was vulnerable to erosion there and the bed structuresmainly consisted of larger particles. This was very different from the upstream river bed where particles ofD< 100 mm still played an important role in structure components. On the other hand, stronger bed structures could supply greater resistance to the river flow. Thus, channels with larger slopes can be sustained since slope has a direct and proportional relationship with stream power. In other words, stronger bed structures (larger scales) can stabilize a larger slope channel. This phenomenon is also shown in Table 3.

Fig. 8 Variation of plane view of channels in bedload scheme

Fig. 9 Variation of 2# cross-section in bedload scheme (phase 1)

3.2 Non-bedload scheme

The non-bedload scheme ran in phase 2 and phase 3 (Table 1). Both of them came after the initial runs of the bedload scheme. Sedimentation tanks were made adjacent to the cement Parshall flume-type flowmeter when the previous run (bedload scheme) had reached equilibrium. Without any alteration in diversion, the bedload could be removed. Sometimes the deposits in tanks had to be cleaned when the experiment was still running, since the bedload transportation rate was very high.

The longitudinal profiles in equilibrium status after scouring of relatively clean water without bedload changed a lot (Fig. 10). A severe incision occurred during this process. In phase 2 the elevation of every location on the thalweg was lowered by 0.2 m to 0.5 m (Fig. 10).Compared with the drastic change in the longitudinal profile, Fig. 11 shows that the channel’s plane view seemed to be stationary. The channel just stopped wandering and started to incise rapidly as soon as the hydrological scheme was switched. Thus, the channel’s horizontal migration could be neglected. After several days a deep main channel was formed and the previous sand bars along the banks became “river terraces” (Fig. 12). That is to say, channels tend to be deeper and narrower when the bedload from upstream was filtered.

Fig. 10 Equilibrium profile in non-bedload scheme

Fig. 11 Variation of plane view of channels in non-bedload scheme

Fig. 12 Variation of 12# cross-section in non-bedload scheme (phase 2)

This kind of incision suggests that the former bed structures evolved in bedload schemes cannot maintain stability after the removal of bedload. Therefore, the river bed was scoured and elevation lowered. The incision depth in the upper reach was relatively larger than the one in the lower reach to some extent since the existing bed structures in the upper reach were weaker (Table 4). Because of this uneven incision, the shape of the longitudinal profile changed from a convex curve to a nearly straight line (Table 4). The bed structures were strengthened especially in the upstream section so that they were gradually similar to the downstream structures. The measurement shows that structures in the upper and lower reaches were almost the same in scale under equilibrium (Table 4). The variation in bed structures also led to the adjustment of the slope they could sustain, which further explains the straight equilibrium profile. As mentioned above, mild-slope sections and steep-slope sections could not be easily distinguished. It can be concluded that the bedload consumes a proportion of stream power and thus is relevant to the slope and the channel profile.

Table 4 Comparison of bed structures and slopes of two schemes (phase 2)

3.3 Artificial bed structures

The non-bedload scheme running in phase 3 was influenced by human interventions (Table 1). After the previous bedload scheme ran, a group of artificial step-pool structures were placed in the middle reach. The artificial structures were about 10 cm in height and50 cmin length (Fig. 13). As noted, in phase 2 the longitudinal profile experienced severe incision in the non-bedload scheme (Fig. 10). In phase 3 human interventions were introduced so that a full-scale incision was impeded (Fig.10). Although the upper reach was still eroded to a certain depth, the middle and the lower reaches maintained their original bed elevations. Therefore, the longitudinal profile in equilibrium status was adjusted by the artificial structures.

Fig. 13 Deployment of artificial step-pool structures

The application of artificial bed structures strengthened the resistance in the middle reach, consuming more stream power than before. Thus, the removal of bedload could not exert too much impact on the channel, especially on the middle reach. Furthermore, this artificial structure section served as a stable base level of erosion that ensured the upper reach would not be incised too much. Thus, to some extent an artificial bed structure here can be regarded as a dam in macro geomorphology. In some reaches bank erosion and sedimentation were detectable, which means that the direction of fluvial process had been changed radically.

Table 5 Effect of artificial bed structures

4 Conclusions

Based on the analyses of the equilibrium profile, bed structure evolution and sediment transportation under different hydrological conditions, the following conclusions can be made:

(1) When the channel is in dynamic equilibrium status, the evolved bed structures have a close relationship with slope. Stronger bed structures can consume more stream power and thus sustain larger slopes. The slope may also be affected by hydrological conditions (discharge and bedload transportation rate).

(2) The bedload can also consume a proportion of stream power. Therefore the slope that certain bed structures can sustain may be influenced by the bedload transportation rate. The structures on the river bed will be eroded if the bedload is removed.

(3) In the natural cases, the distribution of evolved bed structures is related to the environmental conditions (in this experiment only the hydrological conditions were tested). The same environmental condition will result in the same distribution and eventually the same longitudinal profile.

(4) To a certain extent, the artificial bed structures can serve as the natural ones for energy consumption, stream incision resistance, and channel stability.

Abrahams, A. D., Li, G., and Atkinson, J. F. 1995. Step-pool stream: Adjustment to maximum flow resistance.Water Resources Research, 31(10), 2593-2602. [doi:10.1029/95WR01957]

Chin, A. 1999. The morphologic structure of step-pools in mountain streams.Geomorphology, 127(3-4), 191-204. [doi:10.1016/S0169-555X(98)00083-X]

Chin, A. 2002. The periodic nature of step-pool mountain streams.American Journal of Science, 302(2), 144-167. [doi:10.2475/ajs.302.2.144]

Curran, J. H., and Wohl, E. E. 2003. Large woody debris and flow resistance in step-pool channels, CascadeRange, Washington.Geomorphology, 51(1-3), 141-157. [doi:10.1016/S0169-555X(02)00333-1]

Hassan, M. A., and Church, M. 2000. Experiments on surface structure and partial sediment transport on a gravel bed.Water Resources Research, 36(7), 1885-1895. [doi:10.1029/2000WR900055]

Kim, J. S., Lee, C. J., and Kim, W. 2010. Roughness coefficient and its uncertainty in gravel-bed river.Water Science and Engineering, 3(2), 217-232. [doi:10.3882/j.issn.1674-2370.2010.02.010]

Reid, J. B., Jr. 1992. The Owens River as a tiltmeter for Long Valley Caldera, California.The Journal of Geology, 100(3), 353-363. [doi:10.1086/629637]

Rosport, M. 1997. Hydraulics of steep mountain streams.International Journal of Sediment Research, 12(3), 99-108.

Sear, D. A. 1996. Sediment transport processes in riffle-pool sequences.Earth Surface Processes and Landforms, 21(3), 241-262. [doi:10.1002/(SICI)1096-9837(199603)21:3<241::AID-ESP623>3.0.CO;2-1]

Strom, K. B., Papanicolaou, A. N., Billing, B., Ely, L. L., and Hendricks, R. R. 2006. Characterization of particle cluster bedforms in a mountain stream.Proceedings of World Water and Environmental Resources Congress 2005: Impacts of Global Climate Change. Anchorage: American Society of Civil Engineers. [doi:10.1061/ 40792(173)399]

Wang, Z. Y., Xu, J., and Li, C. Z. 2004. Development of step-pool sequence and its effects in resistance and stream bed stability.International Journal of Sediment Research, 19(3), 161-171.

Wang, Z. Y., and Lee, H. W. 2008.Integrated River Management.Beijing: Tsinghua University and the University of Hong Kong.

Wang, Z. Y., Melching, C. S., Duan, X. H., and Yu, G. A. 2009. Ecological and hydraulic studies of step-pool systems.Journal of Hydraulic Engineering, 135(9), 705-717. [doi:10.1061/(ASCE)0733-9429(2009)135: 9(705)]

Whittaker. J. G. 1987. Sediment transport in step-pool streams. Thorne, C. R., Bathurst, J. C., and Hey, R. D., eds.,Sediment Transport in Gravel-Bed Rivers, 545-579. Chichester: John Wiley and Sons.

Wittenberg, L. 2002.Structural Patterns and Bed Stability of Humid Temperate, Mediterranean and Semi-arid Gravel Bed Rivers. Ph. D. Dissertation. Tyne and Wear: University of Newcastle upon Tyne.

Yang, S. F., Hu, J., and Wang, X. K. 2006. Incipient motion of coarse particles in high gradient rivers.International Journal of Sediment Research, 21(3), 220-229.

Yu, G. A., Wang, Z. Y., Zhang, K., and Liu, H. X. 2008. Field experiment for harnessing incised mountain stream (the Diaoga River) by using artificial step-pools.Journal of Hydroelectric Engineering, 27(1), 85-89. (in Chinese)

This work was supported by the National Natural Science Foundation of China (Grant No. 51009096) and the Research Fund of Nanjing Hydraulic Research Institute (Grant No. Y210003).

*Corresponding author (e-mail:liuhx@nhri.cn)

Received Jul. 21, 2010; accepted Jan. 3, 2011