2 Surrogate t echnologies for m onitoring b ed - l oad t ransport in r ivers John R. Gray 1 & Jeffrey W. Gartner 1 (editors) Jonathan S. Barton 2 , Janet Gaskin3 , Smokey A. Pittman 4 & Colin D. Rennie 3 1 United States Geological Survey, USA 2 National Aeronautics and Space Administration, USA 3 University of Ottawa, Canada 4 Graham Matthews and Associates, USA Surrogate technologies for bed - load transport moni- bration, and data analysis), reliability, robustness, toring are being evaluated toward eventually sup- accuracy, size and location of the instantaneous and planting traditional data - collection methods that time - integrated measurement realm, and range in require routine collection of physical samples and size of bed - load particles. Most if not all surrogate subsequent fi eld or laboratory analyses. Commercially technologies for monitoring bed load, including available and prototype technologies based on active - passive and active hydroacoustics, require periodic and passive - hydroacoustic principles are the foci of site - specifi c calibrations to infer transport rates much of the current research on bed - load surrogate occurring over the entire channel cross section. techniques, and are the subjects of this chapter. Should bed - load surrogate technologies prove suc- Field and laboratory tests of bed - load surrogate - cessful in a wide range of applications, the monitor- monitoring techniques using active hydroacoustics ing capability could be unprecedented, providing the (acoustic Doppler current profi lers (ADCPs)) in prospect of obtaining continuous records of bed - load sand - and gravel - bed rivers or passive hydroacoustics discharge potentially qualifi ed by estimates of uncer- (various sensors) in gravel - bed rivers have been tainty. As with suspended - sediment surrogate tech- shown to provide useful data in a limited number of nologies, the potential benefi ts could be enormous, fl ume and fi eld tests, and some are the subject of providing for more frequent and consistent, less continuing research. Research on other technologies expensive, and arguably more accurate bed - load including tracer - tracking (visual, radioactive, mag- data obtained with reduced personal risk for use in netic, and radio); sonar, load - cell, videography, managing the world ’ s sedimentary resources. particle - tracking, ground - penetrating radar, and magnetic techniques is ongoing in several countries. 2.1 Introduction Similar to choices for monitoring suspended - sedi- ment transport, selection of an appropriate technol- Bed load is the part of total - sediment load that is ogy for bed - load transport monitoring usually entails transported by rolling, skipping, or sliding on the an analysis of the advantages and limitations associ- riverbed (ASTM International 1998 ) (Fig. 2.1 ). ated with each technique, the monitoring objective, Historically, bed - load data for US rivers have been and the physical and dynamic sedimentary charac- produced by gradation and gravimetric analyses per- teristics at each deployment site. Some factors that formed on samples obtained with manually deployed may limit or enhance the effi cacy of a surrogate samplers (Edwards & Glysson 1999 ; Kuhnle 2008 ). technology used to monitor bed - load transport As with suspended sediment, traditional bed - load include cost (purchase, installation, operation, cali- data - collection methods tend to be expensive, labor intensive, time - consuming, diffi cult, and under some conditions, hazardous. Specialized instruments Sedimentology of Aqueous Systems, 1st edition. and considerable training in their proper deployment Edited by Cristiano Poleto and Susanne Charlesworth. are prerequisites for obtaining reliable bed - load © 2010 Blackwell Publishing samples. 46 Surrogate technologies for monitoring bed-load transport in rivers 47 Total sediment load By origin By transport By sampling method Suspended load Wash load Suspended load Unsampled load1 Bed-material load Bed load Bed load 1That part of the sediment load that is not collected by the depth-integrating suspended-sediment and pressure-difference bed-load samplers used, depending on the type and size of the sampler(s). Unsampled-zone sediment can occur in one or more of the following categories: (a) sediment that passes under the nozzle of the suspended-sediment sampler when the sampler is touching the streambed Fig. 2.1 Components of total - sediment and no bed-load sampler is used; (b) sediment small enough to pass through the load considered by origin, by transport, bed-load sample’s mesh bag; (c) sediment in transport above the bed-load sampler and by sampling method. that is too large to be sampled reliably by the suspended-sediment sampler; and From Diplas et al. (2008) . (d) material too large to enter the bed-load-sampler nozzle. October 9, 1989 October 11, 1989 October 12, 1989 6.0 3.0 0 Helley–Smith minus BL-86-3 –3.0 –6.0 7.5 Helley–Smith 6.0 BL-86-3 4.5 Bed load (t/day/m) 3.0 1.5 0 10 11 12 13 14 15 16 11 12 13 14 15 16 17 18 Time (h) Fig. 2.2 Variability in sand bed - load transport rates measured 2 meters apart by a Helley – Smith bed - load sampler and a BL - 86 - 3 bed - load sampler (the latter identical to the US BL - 84 bed - load sampler), at the U.S. Geological Survey (USGS) streamgage on the Colorado River above National Canyon near Supai, Arizona, USA, October 1989. From Gray et al . (1991) . The spatiotemporal distribution of bed material apart during steady fl ows near the middle of the sand - transport is a complicated, non - linear function of bedded Colorado River above National Canyon near sediment supply, bed state, and fl uid forcing (Gomez Supai, Arizona, USA (Gray et al. 1991 ). Such variabil- 1991 ). Figure 2.2 shows variations in bed - load trans- ity is more or less typical for at - a - point bed - load port rates measured by two types of pressure - differ- measurements. However, after collection of 390 ence sampler deployed at fi xed locations 2 meters discrete bed - load transport samples using two types 48 Chapter 2 7.5 Upper Whisker 75 percentile 6.0 Median 25th percentile Lower Whisker 4.5 Probable outlier Extreme value 3.0 Bed-load transport rate (t/day/m) 1.5 Left edge of water Right edge of water 0 0 9 18 27 37 46 55 64 73 82 Station cross section (m) Fig. 2.3 Spatially averaged transport rates computed from 390 bed - load samples collected by a Helley – Smith bedload sampler and a BL - 86 - 3 bedload sampler (the latter identical to the US BL- 84 bed - load sampler), at the USGS streamgage on the Colorado River above National Canyon near Supai, Arizona, USA, October 1989. From Gray et al. (1991) . of pressure- difference sampler from points across the methodologies that enable acquisition of temporally channel, a pattern in bed - load transport became evid- and (or) spatially dense fl uvial - sediment data sets ent with most bed load occurring in the center third of without the need for routine collection and analysis the river (Fig. 2.3 ). These data are illustrative of the of physical samples other than for periodic calibra- fact that bed - load data collected by traditional manual tion purposes. Bed - load surrogate technologies have techniques as part of periodic or runoff - initiated site been addressed as part of at least three workshops visits are rarely suffi cient to reliably characterize the held since 2002, namely: spatiotemporal variability in bed - load transport rates • Erosion and Sediment Transport Measurements in over periods exceeding a fraction of a day. Rivers: Technological and Methodological Advances, Lacking a reliable means for developing a bed - load June 19 – 21 2002, Oslo, Norway, convened by the transport time series, practitioners often revert to International Commission of Continental Erosion of estimations based on stochastic techniques, such as a the International Association for Hydrological bed - load transport equation or an empirically derived Sciences, and sponsored by the Norwegian Water bed - load transport curve with instantaneous water Resources and Energy Directorate (Bogen et al. discharge as the independent variable (Glysson 1987 ; 2003 ). Gray and Sim õ es 2008 ). However, the uncertainty • Federal Interagency Sediment Monitoring associated with bed - load - discharge estimates is rarely Instrument and Analysis Research Workshop, quantifi ed or quantifi able, and is more often the September 9 - 11 2003, Flagstaff, Arizona, USA, subject of speculation rather than reliable calculation. sponsored by the Advisory Committee on Water Thus, considerable interest and effort has been Information ’ s Subcommittee on Sedimentation (Gray directed toward surrogate measurements that may 2005 ). potentially provide a bed - load time series that is rep- • International Bedload Surrogate Monitoring resentative of the cross section or reach of interest. Workshop, April 11 - 14 2007, Minneapolis, Sediment - surrogate technologies are defi ned as Minnesota, USA, sponsored by the Advisory instruments coupled with operational and analytical Committee on Water Information ’ s Subcommittee Surrogate technologies for monitoring bed-load transport in rivers 49 on Sedimentation (Gray et al. 2007 ; Laronne et al. nologies, and to related data - management and 2007 ). fl ux - computation issues. This matrix is reproduced The 2002 workshop in Oslo, Norway, included 13 herein as Table 2.2 . About 50 participants from nine papers under the category, “ bed - load monitoring countries attended the 2007 workshop in Minneapolis, and transport processes.