HISTORICAL BACKGROUND During the late 1960's, the Helley-Smith bedload sampler was developed by the California District. Its original design included a 3-inch by 3-inch square entrance and was based on the Arnhem bedload sampler that was developed during the 1930's. These samplers are the pressure-differential type; that is, they have nozzles that have larger cross-section openings at the rear of the nozzle than at the entrance. This can cause the flow to accelerate as it passes into the nozzle (Helley and Smith, 1971). The Arnhem and Helley-Smith samplers have an expansion ratio (ratio of exit area to entrance area) of 3.22. During 1973-76, extensive calibration testing of the Helley-Smith sampler was conducted in the field. Results of these tests were published in Professional Paper 1139 (Emmett, 1980). Based on these results, the Quality of Water Branch (QWB), now the Office of Water Quality distributed QWB Technical Memorandum Nos. 76.04, 77.07, 79.17, 80.07, and WRD Memorandum No. 77.60. These memorandums set interim policy and guidelines for the use of the Helley-Smith sampler, methods for data collection, and publication of the data in the annual data reports.

During the late 1970's, under the auspices of the Technical Committee of the Federal Interagency Sedimentation Subcommittee, extensive flume studies were conducted on various bedload samplers, including the original Helley-Smith sampler and several versions having different nozzle geometries, by the Federal Interagency Sedimentation Project at the St. Anthony Falls Hydraulic Laboratory. On May 15, 1985, as a result of these studies, the Sedimentation Subcommittee accepted a modification of the original 3- by 3-inch Helley-Smith nozzle as the "tentative" standard for use by all Federal agencies. This nozzle has the same entrance and length dimensions as the original 3- by 3-inch nozzle, but it has an expan-sion ratio of 1.40 instead of the original 3.22. It is called the FIASP sampler nozzle. The tentative acceptance was reaffirmed at the April 1988 meeting of the Technical Committee.

During the 1970's and 1980's, considerable research has been conducted on bedload-transport. This research reaffirmed the extreme temporal and spatial variation in the transport rate. Depending on the relative magnitude of the temporal and spatial variability, different sampling methods are optimal when making a bedload-discharge measurement. Although it is important, selection of the proper nozzle expansion ratio, or even sampler type, is probably less critical than the selection of the proper sampling method for obtaining an accurate bedload-discharge measurement.

One of the fundamental features of bedload movement is the extreme variation of the transport rate even when the flow is constant. The theory of dune movement indicates that an individual bedload sample may contain a mass of sediment which ranges from zero to four times that expected from the actual average bedload of the stream (Hubbell, 1987). Field data have shown that both spatial and temporal variability is as large, or larger than predicted by this theory (Emmett, 1980, Carey, 1985, Pitlick, 1988). Thus, an individual bedload sample may give no more than the roughest estimation of the actual average bedload transport. Significant spatial variability also occurs because bedload commonly moves within a narrow portion of the channel width, which is not necessarily coincident with the zone of maximum bed shear stress (Emmett, 1980). Individuals measuring bedload should be well aware of this variability and of the potential inaccuracy of their data.


The following guidelines for the collection, publication, and storage of bedload data supersede previous policy and guideline statements given in QWB Technical Memorandum Nos. 76.04, 77.07, 79.17, and 80.07, as well as WRD Memorandum No. 77.60.

Acceptable Conditions

It is the WRD policy that samples can be collected wherever physical conditions will permit sample collection. Physical conditions will permit sample collection when:

the bed material is firm enough to physically support the sampler without it sinking into the bed;

the streambed is smooth enough for the nozzle to lay flat on the bottom;

the stream velocity is low enough to allow the sampler to properly sit on the streambed (tetherlines allow the sampler to be used in higher velocities than would be possible without them), and

neither organic nor mineral deposits clog the bag to the extent that flow through the sampler is restricted.

Sampler Selection

a. Sampler type

The WRD has made a considerable investment of resources in studying the sampling characteristics of pressure difference type bedload samplers. The WRD is also aware that many other types of bedload samplers are available and are in use throughout the world. In the interest of continuity and consistency, all bedload sampling efforts undertaken by the WRD should use the pressure difference type samplers recommended in this memorandum. Alternative samplers may be considered for sampling in situations outside the range of the recommended type samplers or in situations where funding is available to construct an efficient bedload trap in the streambed. Contact the Office of Surface Water regarding the use of alternative samplers and storage of the data collected.

Schematics of the Helley-Smith, and FIASP samplers are shown in Figure 1. Figure 2 contains a schematic of the wading sampler which may contain either a Helley-Smith or a FIASP nozzle. Working drawings of the samplers can be obtained from the Federal Interagency Project in St. Anthony Falls, Minnesota.

Samplers should be fabricated exactly to design specifications because it has been shown that relatively minor variations in construction can cause significant differences in sampling efficiency. For example, Childers (written communica-tion, December 15, 1989) suggests that certain samplers have different sampling rates, apparently because the frame tubing is not attached at the bottom of the nozzle as specified in the drawings. Likewise, Pitlick (1988) has shown that the sampling efficiency is a function of the wall thickness of the nozzle.

The sampler type must be recorded as part of the basic data associated with any bedload measurement. If a sampler is used for which no parameter code value exists, it should be entered into WATSTORE as type "other" (8010.00) and the full description given in the station analysis.

Figure 1

Figure 1. Schematic of the cable-suspended Helley-Smith (top) and FIASP (bottom) bedload samplers.

Figure 2

Figure 2. Schematic of wading bedload sampler. Either the Helley-Smith or the FIASP nozzle may be used.

b. Bag mesh size:

Field personnel are to use their own judgment in the selection of the proper mesh size. Mesh sizes normally used are 0.25, 0.5, 1.0, and 2.0 mm with the 0.25 mm size being the most common. Large quantities of fine material (about equal in size to the mesh opening) and large organic material (such as leaves) can clog the sampler bag, and invalidate the sample. Increasing the mesh size of the bag may decrease clogging. The mesh size should be selected to minimize both the clogging and the loss of the fine bedload through the mesh.

c. Nozzle expansion ratio:

The WRD endorses the use of the 1.40 expansion ratio nozzle tentatively recommended by the Technical Committee of the Federal Interagency Sedimentation Subcommittee. It is realized that differences between the 1.40 and 3.22 expansion ratio have not been completely quantified but are likely to be small relative to potential errors induced by suboptimal sampling procedures. It is, therefore, acceptable to use samplers with either the 3.22 or 1.40 expansion ratio.

d. Nozzle size:

At present (1990), no established criteria exists for the selection of size nozzle. The nozzle, however, should be at least two times the size of the largest particles likely to be in motion. Some hydrologists recommend the nozzle size be at least 5 times the size of the largest particle in motion.

Sampling Procedure

a. particle-size analysis:

It is mandatory that particle-size analyses of the bedload and suspended load are made. It is not necessary to perform such analyses on every set of samples, although enough analyses should be performed to define the particle size distribution when bedload samples are to be collected at the site. The particle sizes can vary with changes in either flow conditions or season of the year. If at all possible, representative bed material samples should be collected and particle size analyses conducted. The volume of the sample should be sufficient to accurately define the sizes present in the bed.

b. Tetherlines:

The use of tetherlines, often called stay-lines, is strongly recommended when bedload samples are collected using a cable suspended sampler. A tetherline is a line connected between an upstream support, such as a transverse cable, and the sampler. Tetherlines have been used by hydrographers in Europe and elsewhere for many years to eliminate the downstream drift of equipment caused by high-velocity and deep depths in open-channel flow. An article in the July-December 1988 issue of the WRD Bulletin decribes their use and benefits. A tetherline, attached as shown in Figure 1, makes the sampler more controllable and keeps it from drifting downstream while being lowered through the sampling vertical. Tetherlines also help eliminate cross-stream "swimming" of the sampler. The use of tetherlines, when collecting bedload samples using a cable suspension, is believed to reduce the potential for scooping during sampler placement and retrieval.

c. Cross-sectional procedures:

The U.S. Geological Survey Open-File Report 86-531 lists three cross-sectional procedures that can be used. They are the Single Equal Width Increment (SEWI) method, the Multiple Equal Width Increment (MEWI) method, and the Unequal Width Increment (UWI) method. The SEWI method actually involves collecting 2 samples at each of 20 verticals and is the most commonly used procedure. Sampling procedures should consider spatial (cross-sectional) variations and temporal (at-a-point) variations. Because knowledge of where bedload has occurred in the past does not necessarily imply where bedload will occur in the future, it is unlikely that pre-judgment will allow an investigator to substitute spatial concerns for temporal concerns, or vice versa.

It is the responsibility of the field personnel to select the procedure that is optimal for the local conditions. Ordinarily, excess samples will have to be collected at a site until enough experience is gained to select the appropriate procedure for the measurement. At some sites it may be possible to define the temporal and spatial variation with 40 samples and other sites may require 100 or more samples to obtain the same information. A plot of the individual bedload transport rates as a function of distance from the bank has often been found to be useful in selecting the appropriate sampling procedure (Dinehart, personal communication, 1990). The optimal sampling procedure should be expected to vary with season and flow conditions.

Ultrasonic depth soundings can be very useful for evaluating bedload variations (Hubbell, 1964).

d. Sampling time

The sampling time is the length of time the sampler rests on the bed and accumulates material for each individual sample. Long sampling times tend to average out short-term rate fluctuations and produce rates that vary about the mean rate in accordance with the normal distribution. Short times, however, tend to reflect instantaneous rate variations and often correspond to asymmetrical distributions (Gomez et. al., 1989). Because sampler capacity is rela-tively small, the range of acceptable sampling times is usually very limited, often 30 to 60 seconds. Although a sampling time of less than 10 seconds should be avoided, the sample bag should never be filled to more than about half full. The sampling time should not be so long that a significant amount of clogging of the bag occurs.

e. Sample compositing:

A bedload sample is the material collected during one sampling time. The width increment is the width of channel to be represented by the sample vertical (Edwards and Glysson, 1988, p. 98). Individual bedload samples can be (1) analyzed individually, or (2) combined into one or more composite samples for analyses. Only samples collected with equal sampling times and width increments may be composited.

Until the sampling variability for the site is understood, all samples should be analyzed individually. The more samples one composites, the less one learns about variability in space and time.


a. Computation of Bedload Discharge

The usual sampling strategy should be to define the mean transport rate at a series of lateral locations, often equally spaced. For this strategy, samples from different verticals must not be composited. Rather, sample rates should be plotted on a graph of RATE versus WIDTH. The RATE should be expressed in units of weight per unit time per unit width. For conditions of reasonably constant discharge, the rates at each measurement vertical should be averaged for plotting, and a curve representing the lateral distribution of bedload transport rate across the channel (or channels) should be drawn. This curve should give insight into the true shape of the lateral distribution curve. The area under the curve then represents the total cross-section bedload transport rate, in weight per unit time.

b. Sampler efficiency

The efficiency of bedload samplers are generally measured in two different ways. One way is to measure the ratio of the flow of water that passes through the nozzle relative to the amount that would pass through the opening area of the nozzle if the sampler were not there. This is called the hydraulic efficiency. A value of 1.54, for example, indicates that 1.54 times as much flow passes through the sampler as would pass through the opening space of the nozzle if the sampler were not in place.

The more relevant efficiency is the sediment-trapping efficiency, which is the amount of material trapped in the sampler relative to the amount of material that would have passed through the space occu-pied by the sampler nozzle if the sampler were not in place. The sediment-trapping efficiency is much more difficult to determine as indicated in the references.

Bedload data stored in WATSTORE should not be adjusted for sediment-trapping efficiency.

Documentation Bedload data may be stored in WATSTORE and published in the annual data reports. The data will be qualified, however, by storing with each sample data set the mandatory information identified in Table 1. Table 2 contains a partial list of parameter codes that should be used to store bedload data in WATSTORE.

Table 1. Mandatory and Optional Data to be Stored with Bedload Samples

Mandatory data

Parameter Code
Date (begin date, end date)
Time of compostie (begin time, and end time)
Hydrologic condition code
Start Time, of measurement
End time, of measurement
Instantaneous stream discharge (cfs)
Bedload discharge total section (tons/day)

Composites in section discharge (count)
Short name - Number of composites

Bedload transport rate, unit (tons/day/ft)
Short name - Bedload transport rate
Number of samples in composite (count)

Number of verticals in comosite
Short name - Number of verticals

Cross-section location feet from left bank (ft)
(Vertical in composite closest to the left bank)
Width increment of samples in composite (ft)
Short name - width increment of sample
Sampling time of each bedload sample (sec)
Short name - Sampling time of sample
Type of samplet (see Table 2)
Sampling methodology (see Table 2)
Bedload sampler bag mesh size (mm)
Tetherline (0=no, 1=yes)
Short name - Tetherline (0=no, 1=yes)
Bedload particle size distribution
Suspended sediment concentration (mg/l)
Suspended sediment particle-size distribution
Bed material particle-size distribution
Stream temperature (deg C)

Optional Data

Hydrologic event
00065 Gage height (ft)
00055 Stream velocity (ft/s)
00004 Stream width (ft)
00064 Depth of stream, mean (ft)
01351 Streamflow (severity)
71999 Sample purpose (codes)

Table 1 (continued) Footnotes to Table 1

? Parameter code has been applied for.
(1) Enter the 'Date of measurement' which may be composed of one or more composite samples. If the measurement spans 2 or more days enter the date the measurement was started.

(2) A composite is a sample of bedload material which is bagged for analysis. A composite sample may contain 1 or more individual samples. An individual sample is the material collected while the sampler is on the bed for one sample time. The 'Time of composite' should begin at the time of the collection of the first individual sample in the composite and end with the time of completion of the last individual sample in the composite. The time must be unique for each composite.

(3) The 'Bedload discharge total section' is the total cross sectional bedload discharge as expressed in tons per day.

(4) The 'Composites in section discharge' is the number of composited samples (one or more) making up the total cross sectional measurement.

(5) The 'Bedload transport rate' is the average unit bedload discharge for the samples in the composite expressed in tons per day per foot of width.

(6) A composite sample may contain one or more individual samples obtained at one or more verticals. The 'Number of samples in composite' is the total number of individual samples combined in the composite sample.

(7) The 'Number of verticals in composite' is the number of verticals, one or more, represented in the composite sample.

(8) The composite sample may contain individual samples from several verticals. The 'Cross-section location' should represent the distance from the left bank to the closest vertical included in the composite. Only contiguous verticals may be composited.

(9) The 'Width increment of samples in composite' is the horizontal width of of river represented by each vertical in the composite. Only verticals with equal width increments may be composited.

(10) The 'Sampling time of each bedload sample' is the actual time in seconds that the sampler rests on the bed for each individual sample. Only samples with equal sampling times may be composited.

(11) The code 'Tetherline' will indicate whether or not a tetherline was used for collecting the samples in the composite.

(12) It is not necessary to run a size analysis for every composite sample but enough analyses should be performed to accurately define these parameters for all seasons and flow conditions in which bedload samples are to be collected at the site.

Table 2. Values for selected Parameter Codes

Sampling methodology, Code 82398

Value Description
SEWI, Single Equal Width Increment method
MEWI, Multiple Equal Width Increment method
UWI, Unequal Width Increment method
Single vertical

Type of sampler, Code 84164

3" by 3" Helley-Smith, 1/4-inch thick nozzle and tail, Weight (50 to 100 pounds), cable suspended, Expansion ratio 3.22.
3" by 3" Helley-Smith, 1/4-inch thick nozzle and tail, Weight (100-200 pounds), cable suspended, Expansion ratio 3.22.
3" by 3" Helley-Smith, 1/4-inch thick nozzle, no tail, wading rod suspended, Expansion ratio 3.22.
3" by 3" Helley-Smith, 16-gage nozzle, no tail, wading rod suspended, Expansion ratio 3.22.
3" by 3" FIASP, 1/4-inch thick nozzle and tail, Weight (50-100 pounds), cable suspended, Expansion ratio 1.40.
3" by 3" FIASP, 1/4-inch thick nozzle, no tail, wading rod suspended, Expansion ratio 1.40.
3" by 3" FIASP, 16-gage nozzle, no tail, wading rod suspended, Expansion ratio 1.40.
6" by 6" Helley-Smith, 1/4-inch thick nozzle and tail, Weight (150-200 pounds), cable suspended, Expansion ratio 3.22.
12" by 6" Hubbell #5, 1/4-inch plate nozzle and tail, Weight 100-150 pounds, cable suspended, Expansion ratio 1.40.
6" by 12" Toutle River type 2, 1/4-inch plate nozzle and tail, Weight (100-200) pounds, cable suspended, Expansion ratio 1.40.

Districts should review bedload data stored in WATSTORE prior to the issuance of this memorandum and enter as much of the mandatory and optional information as possible.

Questions and/or comments concerning this new policy should be directed to the Chief, Office of Surface Water, 415 National Center, Reston, Virginia 22092.

SUPPLEMENTAL INFORMATION Because of spatial and temporal variations in bedload, and the large effort required to obtain even a few samples, the measurement of bedload is substantially different than the measurement of streamflow and suspended-sediment dis-charge. In particular, it is not possible to establish rigid procedures or accuracy criteria to insure suitable results. Satisfactory measurements can be obtained only by sampling according to an individually planned effort designed to account for the unique situations at a specific measurement site. Optimal schemes require some understanding of the phenomena of bedload transport, particularly with regard to rate variations. Accordingly, it is recommended that anyone who programs, plans, or executes bedload- sampling efforts should seek the wisdom of experienced bedload- sampling personnel and become familiar with references in the literature, especially those given in this policy statement.


Carey, W.P., 1985, Variability in measured bedload-transport rates: Water Resources Bulletin, American Water Resources Association, v. 21, no. 1, p. 39-48.

Childers, Dallas, 1989, Field Comparison of four pressure- difference bedload samplers in high-energy flow: U.S. Geological Survey Water Resources Investigations, in preparation.

de Vries, M., 1973, On measuring discharge and sediment transport in rivers: International Seminar on Hydraulics of Alluvial Streams, New Delhi, India, January 15-19, 1973, Delft Hydraulics Laboratory Publication 106, 15 p.

Druffel, L., Emmett, W.W., Schneider, V.R., and Skinner, J.V., 1976, Laboratory hydraulic calibration of the Helley-Smith bedload sediment sampler: U.S. Geological Survey Open-File Report 76-752, 63 p.

Edwards, T. K., and Glysson, G.D., 1988, Field methods for measurement of fluvial sediment: U.S, Geological Survey Open- File Report 86-531, Reston, Virginia, 118 pages.

Emmett, W.W., 1980, A field calibration of the sediment-trapping characteristics of the Helley-Smith bedload sampler: U.S. Geological Survey Professional Paper 1139, 44 p.

---- 1981, Bedload sampling in rivers: Beijing, China, Proceedings, International Symposium on River Sedimentation, March 24-28, 1980, Guanghua Press, p. 991-1014.

---- 1981, Measurement of bedload in rivers: Florence Italy, Proceedings, International Symposium on the Measurement of Erosion and Sediment Transport, June 22-25, Publication no. 133, International Association Hydrological Sciences, p. 3-15.

---- 1984, Measurement of bedload in rivers, Chapter 5 in Hadley, R.F., and Walling, D.E., (eds.), Erosion and Sediment Yield: Norwich, Geo Books, p. 91-110.

Emmett, W.W, Leopold, L.B., and Myrick, R.M., 1983, Some characteristics of fluvial processes in rivers: Nanjing, China, Proceedings, Second International Symposium on River Sedimentation, October 11-16, Water Resources and Electric Power Press, p. 730-754.

Gomez, Basil, Naff, R.L., and Hubbell, D.W., 1989, Temporal variations in bedload transport rates associated with the migration of bed forms: Earth Surface Processes and Landforms, v. 14, p. 135-156.

Guy, H.P., and Norman, V.W., 1970, Field methods for measurement of fluvial sediment: Techniques of Water-Resources Investigations of the U.S. Geological Survey, Book 3, Chapter C2, 59 p.

Helley, E.J., and Smith, W., 1971, Development and calibration of a pressure-difference bedload sampler: U.S. Geological Survey Open-File Report 8037-01, 18 p.

Hubbell, D.W., 1964, Apparatus and techniques for measuring bedload: U.S. Geological Survey Water-Supply Paper 1748, 74 p.

---- 1987, Bedload sampling and analysis, in Sediment Transport in Gravel-Bed Rivers: Edited by C.R. Thorne, J.C. Bathurst, and R.D. Hey, John Wiley & Sons, New York. Hubbell, D.W., and Stevens, H.H. Jr., 1986, Factors affecting accuracy of bedload sampling: Proceedings of the Fourth Federal Interagency Sedimentation Conference, v. 1, p. 1-20 to 4-29.

---- 1986, Laboratory data on coarse sediment transport for bedload sampler calibration: U.S. Geological Survey Water- Supply Paper 2299, 31 p.

Hubbell, D.W., Stevens, H H., Jr., Skinner, J.V., and Beverage, J.P., 1985, New approach to calibrating bedload sampler: Journal of Hydraulic Division, American Society of Civil Engineers, v. 111, no. 4, p. 677-694.

---- 1986, Videotape: Characteristics and use of Helley-Smith type bedload samplers: U.S. Geological Survey Open-File Report, 25 minutes.

McLean, D.G., Tassone, B., 1987, Discussion of bedload sampling and analysis, by D.W. Hubbell, in Sediment Transport in Gravel- Bed Rivers: Edited by C.R. Thorne, J.C. Bathurst, and R.D. Hey, John Wiley & Sons, New York, p. 109-114.

Pitlick, J., 1987, Discussion of bedload sampling and analysis, by D.W. Hubbell, in Gravel-Bed Rivers: Edited by C.R. Thorne, J.C. Bathurst, and R.D. Hey, John Wiley & Sons, New York, p. 106- 108.

---- 1988, Variability in bedload measurement: Water Resources Research, v. 24, no. 1, p. 173-177.