Limnol. Oceanogr., 44(7), 1999, 1793–1801

The acoustic Doppler velocimeter (ADV) has recently been suggested as a promising instrument for characterizing near-bed flows, particularly in the first 10 mm above the bed where many benthic organisms live. Flow characteristics in such settings often exhibit steep vertical gradients, however, so the reliable use of the ADV requires knowledge of the size and location of the instrument’s acoustic sampling volume. We describe simple procedures for quantifying the vertical size of the ADV’s sampling volume and assessing its height above the bed. Our results indicate that the vertical size of the sample volume for the ADV we tested was much larger than expected based on the values predicted by software configuration. Moreover, this system incorrectly reported several distances needed to accurately position the sample volume near the bed (e.g., the transmitter-to-bed distance, the sample volume-to-bed distance, and the transmitter-to-sample volume distance). We also demonstrate that incorrect assumptions about the size and location of the sampling volume can lead to inaccurate near-bed flow measurements by comparing the time-averaged flow speed profiles generated by our ADV with those obtained using a hot-film velocimeter (HFV). At heights . 10 mm above the bed, both instruments yielded similar flow speed estimates. Closer to the bed, however, the flow speeds reported by ADV were as much as 60–80% less than those from HFV. These large errors in estimating near-bed flow speeds are a direct consequence of incorrect assumptions about the centerpoint and size of the ADV’s sample volume. Specifically, when the vertical size of the sampling volume is larger than its nominal size, users may mistakenly position the ADV so that the bed is included within the sampling volume, which in turn results in the underestimation of flow speeds. By validating the size and location of the sampling volume, as well as carefully monitoring signal quality parameters, users can ensure proper placement of the ADV relative to the bed and avoid erroneous measurements. Aquatic scientists working in both marine and freshwater environments often need to measure the near-bed flow conditions that affect benthic organisms and processes (Nowell and Jumars 1984; Davis and Barmuta 1989; Eckman et al. 1990; Dodds et al. 1996). Investigators face several challenges in making accurate and informative measurements of near-bed flow characteristics, however. First, many benthic organisms are relatively small, projecting , 10 mm above the bed. Second, flow fields in this near-bed region are often strongly sheared and are usually three-dimensional among the roughness elements where benthic organisms often live (e.g., on rocky shores or streambeds) (Denny 1988; Carling 1992). Third, and following from the second, it is often difficult to predict near-bed flow characteristics at a specific locale from coarse-scale measurements made farther above the bed (e.g., Hart et al. 1996). Collectively, these factors place a premium on the use of flow instruments with sufficient spatial resolution to make measurements within the steep gradients associated with near-bed flows. Unfortunately, flow measurement devices with high temporal and spatial resolution, such as hot-film anemometry or laser Doppler velocimetry, are often costly or difficult to deploy in the field. The ADV potentially offers solutions to many of the problems associated with near-bed flow measurement. The ADV operates by emitting a burst of sound energy of known duration and frequency that is subsequently reflected back to the probe by suspended particles moving with the water current. Because the suspended material in the water is moving, the backscattered sound energy is shifted in frequency (known as a Doppler shift), and the magnitude of this shift is proportional to flow speed (for more information, see Kraus et al. 1994; Lohrmann et al. 1994; Zedel et al. 1996). ADV systems are sufficiently robust to be field portable, offer moderate temporal and spatial resolution, and do not require routine calibration. Because the ADV measures velocities from a volume of water (or ‘‘sample volume’’) located some distance from the physical probe, ADV measurements are noninvasive, which provides a considerable advantage over hot-film, propeller, or electromagnetic flow sensors. SonTek manufactures the ADV most commonly used in aquatic research. These probes consist of a central circular transmitter (6-mm diameter) surrounded by three equally spaced receivers (Fig. 1). Depending on configuration, the sample volume is located at the intersection of the transmit and receive ‘‘beams,’’ either 50 or 100 mm from the probe (Fig. 1). In addition, the manufacturer specifies that the sample volume for this probe be approximated by a cylinder with a diameter equal to the transmitter diameter (6 mm) and a software configurable height ranging from 1.2 to 9.0 mm (Lohrmann et al. 1994; SonTek 1997). Because the sample volume cannot be directly observed, however, its precise size and spatial position are difficult to define. Accurate knowledge of the size and position of the ADV’s sample volume is essential for near-bed flow measurements. These two parameters not only determine where flow characteristics are measured, but also how close to the bed the transmitter can be placed without the sample volume incorporating part of the bed. For example, the position of the sample volume, and thus the location at which velocities are measured, is referenced to the vertical center of the cylindrical sample volume. Therefore, a measurement made at a height of ‘‘10 mm’’ above the bed is actually a spatial average that is vertically centered at 10 mm. Theoretically, for use near solid surfaces, the ADV can be operated reliably even when the lower end (or ‘‘bottom’’) of the sample volume is placed 0.5 mm above the bed surface (Lohrmann et al. 1994; SonTek 1997). If these assumptions are correct, then the ADV can be safely deployed with the center of the