Geodesy

Physical oceanography

Climatology

Statistics

Ocean current

Calibration

Circulation (fluid dynamics)

Geology

Current (fluid)

Mooring

Oceanography

Thermodynamics

Physics

Climate change

Mathematics

Astronomy

Geography

Ocean observations

Acoustic Doppler current profiler

Remote sensing

General Circulation Model

Meteorology

Cruise

Doppler effect

Journal of Geophysical Research Atmospheres

[1] In a recent paper, Wang et al. [2004] discuss the summer circulation in the South China Sea (SCS) using observations and a generalized topography-following ocean model. Their paper appears to be important since it is based on the only intensive observation with basin-scale coverage of the SCS in late summer. However, their results from underway observations of acoustic Doppler current profiler (ADCP) are not properly processed and may lead to misunderstanding of summer circulations of the SCS. As principal investigators of the research project of the 2000 summer cruise, we would like to present our corrections to the ADCP measurements, which result in more convincing circulation fields, and to call for attentions of properly calibration while using shipboard ADCP measurements. [2] During the cruise, along-track current measurements were conducted using an RDI 150 kHz narrow band vessel-mounted acoustic Doppler current profiler (ADCP). Navigation information was provided to an RDI DAS PC by the ship's gyrocompass and a shipboard commercial GPS for in situ data acquisition. [3] The ADCP velocity vectors at depths of both 100 and 200 m presented in Figure 5 of Wang et al. [2004] look suspicious. If comparing them with ship tracks shown in Figure 1, one can easily see that almost all velocity vectors plotted are pointing toward the port side of the research vessel. This phenomenon suggests that there had been a misalignment of the transducer with respect to the ship's centerline that caused wrong projection of the ship's velocity on the ADCP and led to serious systematic error in the measured currents. Therefore additional calibration after the cruise is necessary [Joyce, 1989; Pollard and Read, 1989]. It is doubtful that such calibration had been included in Figure 5 of Wang et al. [2004]. [4] There are basically two types of errors in shipboard ADCP measurements: an alignment error in the Doppler transducer/ship gyro relative to the “true” heading (the misalignment angle α) and a sensitivity error of the acoustic system factor (the scaling factor 1 + β), where β is relatively small but α could cause severe problems to the measurements [Joyce, 1989; Pollard and Read, 1989]. As the major error source, an error in transducer alignment and ship gyrocompass (α) will produce an apparent athwartship velocity [Joyce et al., 1986] and normally cause a systematical “offset” error sideward of the ship track. [5] On the basis of physics given by Joyce [1989], Pollard and Read [1989] and Joyce et al. [1986], we calibrated the same ADCP data set used by Wang et al. [2004] accordingly. Using all available bottom tracking ADCP measurements from shallow sectors of the cruise, we first estimated these calibration coefficients with bottom tracking method developed by Joyce [1989], which yields a misalignment angle of 2.34° and a scaling factor of 1.0064. Since R/V Xiangyanghong14 was steaming at 10 knots generally, measurement error raised by the misalignment is 0.20 m s−1 roughly in the cross-track direction, which is much greater than that caused by the scaling factor (roughly 0.03 m s−1). [6] Calibrations were also applied for water track mode, because that bottom tracking was not available most of the time for the cruise. In the following analysis, we assume that drifting errors derived from the gyrocompass are random in the 40-day-long cruise; hence that misalignment are basically induced by installation alignment. [9] Starting from the north and east (precalibration) components of Doppler velocity (u′d, v′d)DAS, the vessel's concurrent heading (θ), and the ship speed (U) of DAS output, we first transform these measurements into the localized earth coordinate frame with y axis pointing towards the ship's heading, which yields (u′d, v′d), and vs in the new frame. The misalignment angle (α) and scaling factor (1 + β) were then estimated by least squares linear regression using equations (3′) and (4′), respectively. [10] In the present case, the ship sailed along a zigzag track with speeds ranging from zero (while taking hydrographic stations) to about ten knots (while cruising), a good estimate of α is thus reached. By using ADCP measurements of (u′d, v′d)DAS at the 100 m level, we find a misalignment angle of 2.17° and a scaling factor of 1.0036 from the water track mode. These results are very close to those from bottom tracking and the whole ADCP data set is calibrated using parameters estimated from the water track mode. [11] The calibrated velocity fields at depths of 100 and 200 m are given in Figure 2. The improvement is obvious compared with Figure 5 of Wang et al. [2004]. First, currents at 100 m are generally greater than those at 200 m in a factor of 2, revealing strong baroclinity of the SCS upper circulation not presented before the calibration. Second, the calibrated fields are in better agreement with sea surface height anomaly (SSHA) derived from satellite altimetry [see Wang et al., 2004, Figure 4]. The anticyclonic eddy (W1) east to the Vietnam coast are very well organized in both layers, and a strong offshore jet directing eastward is clearly resolved at 12°N. Figure 2 appears to be the first evidence of the jet from direct current observation, though it had been suggested decades ago by Wyrtki [1961], and latterly seen in altimetry [Shaw et al., 1999; Ho et al., 2000; Li et al., 2003] and AVHRR observations [Kuo et al., 2000]. The jet is obviously associated with the anticyclonic eddy. A northwestward return flow north of eddies C2 and C1 suggested by SSHA [see Wang et al., 2004, Figure 4] is also revealed. This flow manifests a cyclonic circulation pattern north of W1 though measurements in the central SCS are sparking. [12] In summary, shipboard ADCP measurement is very sensitive to misalignment of its installation. In our case, an error of about 0.20 m s−1 sideward to the ship had been generated because of a misalignment angle of 2.17° only. Hence reexamination of shipboard ADCP measurements after the cruise is necessary, and proper calibration may improve results of the observation significantly. This measure is especially important when using ADCP observations from another party. [13] This work was supported by National Basic Research Program of China (grant G1999043807) and the fields cruise by the same program (grants G1999043805, G1999043806, and G1999043807). The cruise is conduced jointly by research groups from The Second Institute and The Third Institute of Oceanography, State Oceanic Administration of China, and The South China Sea Institute of Oceanology, Chinese Academy of Sciences. Special thanks are given to Dongxiao Wang and Jianping Xu for contributions to planning and organizing of the cruise, to Zhaofang Zhu and all scientists onboard for endeavors in the field, and to the captain and crewmembers of R/V Xiangyanghong14. The authors also wish to acknowledge valuable comments by Quanan Zheng, Janping Gan, and an anonymous reviewer.

Comment on “Circulation in the South China Sea during summer 2000 as obtained from observations and a generalized topography‐following ocean model” by H. Wang et al.