Using an ADP from a Moving Boat

1. Introduction

When performing water-current surveys covering large areas, or when monitoring river discharge, it is often convenient to use a boat-mounted SonTek/YSI Acoustic Doppler Profiler (ADP). When operating from a moving platform, an ADP measures relative currents (Figure 1). As such, it is important to measure independently the speed of the platform so it can be subtracted from the raw current measurements. This allows you to obtain the residual water currents relative to the fixed Earth. It is generally desirable to perform these calculations in real-time.

Figure 1. Measuring currents from a moving boat using an ADP

Though there are a several ways to measure a vessel's speed and direction, the two most-practical methods commonly used with ADPs are:

• The use of GPS equipment
• The use of the SonTek bottom-tracking algorithm

2. Using an ADP with GPS

The majority of vessels used for research have GPS systems as part of their navigation equipment. SonTek ADPs manufactured since 1996 have a built-in interface to receive GPS information, and are able to accept the NMEA 0183 data formats (GPGGA/GPGXP/GPGGK) available from most commercial GPS receivers. A good-quality GPS receiver with accurate differential corrections is recommended for robust real-time operations. The GPS-to-ADP interface requires two serial ports (one for the ADP and one for the GPS) on the PC running SonTek's real-time ADP software. The synchronization of data is done in the PC (Figure 2). To compute the vessel's velocity vector, the ADP records GPS positions at the beginning and end of a user-selected averaging interval. The resulting boat velocity is then computed from the total vessel displacement and is subtracted from the relative water currents measured by the ADP. The GPS method for removal of the boat track from ADP measurements is best suited for areas where the bottom is out of ADP tracking range or where currents change slowly in time/space so that longer averaging intervals can be used.

Figure 2. Operating ADP from a boat: connecting necessary equipment

3. SonTek Bottom-Tracking Algorithm

Using the ADP's bottom-tracking algorithm greatly enhances the instrument's capabilities and versatility when used from a moving boat in shallow waters. Bottom-tracking enables you to obtain real-time vessel-speed-over-ground data simultaneously with water current measurements without using any additional equipment. While bottom-tracking, the ADP measures the Doppler shift of reflected acoustic energy (from the bottom of a river, harbor, etc.) to infer the vessel speed. In contrast, when an ADP is current-profiling, it is measuring acoustic reflections from suspended material in the water column to determine the velocity of the water with respect to the ADP. In bottom-tracking mode, the ADP determines bottom velocity once every second. At the end of the averaging interval, all these bottom-velocity estimates are averaged and stored with the profile data. In addition to the vessel-speed-over-ground data, the bottom-tracking ADP also reports an averaged depth in real-time, which is often required for surveys or river discharge applications.

The ADP can only track the bottom when it is within the acoustic range of the ADP’s transducers. Generally, it is good practice to understand the bottom characteristics of the area being surveyed, as some soft bottoms with large amounts of plant growth can imitate "moving-bottom" characteristics that might influence the ADP's bottom-track algorithm. Seabeds with strong acoustic reflections (rocky bottoms provide the best results) can typically be tracked at ranges that are 20-30% greater than the water-current profiling range.

When operating from a moving platform such as a boat, erratic course changes during the averaging interval may yield erroneous vessel velocity vectors. Therefore, when using ADP with GPS or bottom-track, you should take into consideration the vessel's proposed track-line when choosing the averaging interval. The best results will be obtained when steering a vessel in a straight line at a steady speed.

4. GPS vs. Bottom-Tracking

4.1. Performance Comparison

What are the advantages of bottom-tracking over GPS (and vice versa), and what factors affect the inherent accuracy of both approaches? The purpose of both bottom-tracking and GPS is to calculate vessel speed in order to correct the relative water currents measured by the ADP. Therefore, the accuracy of the technique used for vessel speed estimation directly determines the accuracy of the absolute current-velocity profile.

First, consider the pros and cons of the GPS approach. Regular GPS is accurate to within ~100 m, which would require extremely long averaging times (~2.5 h) to achieve a 1 cm/s accuracy. Such measurements are useful only in the open ocean. Differential GPS (DGPS) receivers are capable of providing higher positioning accuracy -- 2-m in the open ocean or even better when in the vicinity of the reference station on land.

Even the most advanced DGPS systems determine vessel position with some uncertainty , which introduces an error into speed calculations:

where is the time between successive GPS fixes (averaging interval).

If the positioning data has an uncertainty of ±2 m, and an averaging interval of 60 seconds is chosen, the uncertainty of the vessel speed for each profile is about ±4.7 cm/s. This uncertainty puts a lower limit on the accuracy of the absolute current measurements. If a current-profiling precision of 1 cm/s is needed, the use of the GPS would require longer averaging intervals and thus limit the time resolution of profiling surveys. Although this may not pose a problem in the open ocean (where currents usually do not change over several minutes), coastal, frontal, and convergent areas require shorter averages. A sub-meter accuracy can be achieved by the latest DGPS systems, however it is usually limited to operations close to a reference station (~100 km), and is not available for remote inland or coastal areas. Also, operating in deep canyons or in mountainous area can cause "shadow," where the direct path to a GPS satellite is obscured.

If DGPS mode is not available during the deployment, necessary corrections can be done in post-processing. ADP software stores the raw GPS positions used for the vessel speed calculations, and differential corrections can be downloaded later from appropriate sources. These corrected positions can be used to obtain more-accurate estimates of the vessel speed, and hence, the true water currents.

Now let us consider the performance of the bottom-tracking algorithm. As mentioned above, a bottom-tracking ADP determines the bottom velocity once each second, and then averages the raw estimates over the user-selected averaging interval. Because the bottom velocity is derived from solid-object reflections, natural variability (standard deviation) of the bottom-track velocity measurements is lower by an order of magnitude when compared to the current-profiling data. Hence, the precision of bottom-velocity measurements is always better than that of water currents. Because of this, bottom-tracking can be considered to introduce no additional errors to water current measurements.

A basic limitation of bottom-tracking is that it can operate in waters with depth no more than approximately (1.3 × Maximum ADP Profiling Range). In addition, significant plant growth, heavy wave conditions, and a moving near-bottom layer can degrade bottom-track performance.

4.2. Compass Alignment

Another important consideration when operating from a moving platform is the accuracy and alignment of the compasses used for obtaining vessel speed and the ADP compass. When using GPS, the ship’s gyro is normally used to provide vessel direction. If a small misalignment (offset) exists between the gyro and the ADP, an error proportional to the speed of the vessel is introduced. That is, for a 5° offset, a vessel moving at 2.5 m/s (5 knots) will introduce a 22 cm/s additional velocity component into ADP measurements.

In the case of bottom-tracking, compass offset will still produce an error in water-current measurements, but this error is proportional to the speed of the current. In most applications, water-current velocities are much smaller than vessel speed, so the introduced errors are smaller. For the same compass offset shown in the above GPS example, and a current speed of 0.5 m/s, the corresponding error is only 4 cm/s.

4.3. Using GPS while Bottom-Tracking with an ADP

To avoid compass misalignment errors, the most robust solution is to use bottom-tracking data together with GPS data. At the beginning of the transect, you can simply compare the ship’s direction derived from bottom-tracking with the gyro output, and then correct for possible misalignment.

The ADP is capable of recording GPS data while simultaneously bottom-tracking. This not only provides a check between the two navigational methods, but it allows the use of GPS when the bottom is out of range. All the advantages of both methods can be used, and the ADP software easily works with both data sets. If both the GPS and bottom-track velocities are collected, you are able to extract water currents relative to either bottom track or to GPS simply by using the software provided with the ADP.

5. Field Test of Bottom-Track Performance vs. GPS

Test data were collected in Mission Bay, San Diego with a 0.25-MHz ADP in January, 1999 (Figure 3). Bottom depth ranged from 80 m at the beginning to less than 10 m at the end of the transect. Ocean swell was less than 0.5 m, and pitch and roll varied within 6°. Although transecting over a bottom of different composition and slope, 99% of the profiles produced percent-good pings greater than 80 (percent-good pings are that portion of the bottom-tracking pings within the averaging interval when the bottom velocity was measured successfully). This confirms the robustness of the SonTek bottom-tracking algorithms. Overall, test results demonstrate a one-percent or better agreement between the GPS and the bottom-track estimates.

The capability of using an ADP from a moving platform greatly extends the versatility of this instrument. Whether you use DGPS, bottom-tracking, or both, you can be assured that the SonTek ADP maintains high standards for quality, value, and reliability.

Figure 3. Field test of bottom-track vs. GPS

6. Results of Bottom-Tracking Test at Offshore Model Basin

Bottom-tracking accuracy of a SonTek Doppler system was tested at the Offshore Model Basin in Escondido, California in December 1999. This facility features a computer-controlled carriage that can move along the basin at speeds up to 5.5 m/s.

The SonTek Doppler instrument was installed in the middle of the carriage with its head submerged about 0.3 m into the water. Two independent runs were conducted at 11 speeds ranging from -3 m/s to 3 m/s. The results of the test are shown below. A least-squares linear fit of the velocity data to the carriage speed gives a slope of 0.9992 with a reported offset of 0.13 cm/s.

The results from this test indicate a bottom-tracking accuracy of 0.08 percent, which is well within our product specification of 0.2 percent.

• ADP