top_home

Data quality of ADCMs

In order to evaluate the performance of a SonTek Argonaut-MD acoustic Doppler current meter (ADCM) mounted on the TRITON buoy, the ADCM was compared with a separately moored upward looking RDI acoustic Doppler current profiler (ADCP) from March 16 to May 17, 1998 (63 days) within a distance of approximately 10 km (exact locations: TRITON buoy 0-5.3 S, 156-1.6 E; ADCP mooring 0-0.03 N, 156-0.04 E). The ADCM was mounted on the TRITON buoy at 197 m depth, where the equatorial undercurrent flows steadily and where spatial differences are expected to be small, and the ADCM data were compared with the ADCP data at the same depth. The amplitude and direction of horizontal current velocity show good agreement between the ADCM and ADCP. For 25-hour running mean data, the correlation coefficient and root-mean-square error of the amplitude were 0.89 and 3.3 cm/sec, and for direction were 0.99 and 3.1 degree, respectively. The mean differences (ADCM minus ADCP) were small, -0.1 cm/sec and -1.4 degree, and the fitting slopes for linear regressions for the amplitude and direction were 1.00 and 1.01, respectively. The mean differences are within manufacturer's nominal accuracy for the ADCM. The ADCM performance is adequate to measure the equatorial currents.


Figure 1   Scatter diagrams of the amplitude (left) and direction (right) between the hourly ADCM and ADCP data. Bold line indicates the linear regression line by the least square fitting forced through the origin.


Figure 2   Time series of the hourly amplitude (lower) and direction (upper) between the ADCM (black line) and ADCP (gray line).


Figure 3   Scatter diagrams of the amplitude (left) and direction (right) between the 25-hour running mean ADCM and ADCP data. Bold line indicates the linear regression line by the least square fitting forced through the origin.