Improved Current Estimates from Spar Buoy-Mounted ADCP Measurement Station: A Case Study in the Ligurian Sea

Round 1

Reviewer 1 Report

Thanks for the nice paper. Only question I have is the design of the experiment. The paper use the historical data and numerical model output for data-data comparison. While it is good, ideally parallel deployment of ADCP mounted in a different platform should be used to calibrate the observation and algorithm.

Author Response

REVIEWER: Thanks for the nice paper. Only question I have is the design of the experiment. The paper use the historical data and numerical model output for data-data comparison. While it is good, ideally parallel deployment of ADCP mounted in a different platform should be used to calibrate the observation and algorithm.

AUTHORS: We fully agree with the reviewer that the comparison with concurrent co-located ADCP measurements from fixed mooring should be the most appropriate strategy for this kind of calibration. Unfortunately, in the lack of a dedicated experiment and funding for an additional mooring building and deployment, we have had to rely on historical and modelling data. Nevertheless, a future set up of a comprehensive experiment with the parallel deployment of another ADCP is among our next objectives and further work along these lines will be hopefully devoted soon.

Reviewer 2 Report

This publication describes the error and uncertainty analysis made on the data of a stand-alone current-meter placed on the damping disk of an ODAS buoy. It shows the important compass error that can be committed by metal parts of the buoy and the care that must be taken in the detection and correction of errors to obtain reliable ocean current data.

Line42: why speaking of ‘years after 2000’, so that in the 80^th and 90^th, there was already RAFOS floats and rotor current-meters on mooring lines to measure currents? In this sentence, a distinction could be made between Eulerian and Lagrangian techniques for currents measurements. Each have advantages and inconvenient.

Figure 1: the location of the buoy is difficult to see on the map.

Line 135: how have you found the uncertainties of 1.5 and 1.7 cm/s? 6.7 * 0.01 = 0.067 cm/s

Line161: how have you found the uncertainty of 0.20 dbar? Nortek pressure sensors are given to have an accuracy of 0.25 % of the full scale. The ADCP is a 200 m. Therefore, 200 * 0.25% = 0.5 dbar or 0.25 dbar if considering the specification as an expanded uncertainty.

Lines 167-169 are repeated two times.

Lines 252-253: could the 14 ° error be due to the positioning error of the ADCP by the diver, because it seems that the average error of figure 10 is close to 0 °?

Line 253: the battery packs can have also a strong influence on the magnetic field.

Line 273: how is made the heading correction, point by point with the errors displayed on figure 10, or by using the average error?

Table 3: could you remember in the legend the significations of letters a, bc and O?

Line 353-354: it must be an error in the reference numbers. Won’t it be [16] and [17]?

Line 365: how are calculated these standard deviations?

Lines 388 and 391: the reference numbers are erroneous.

Author Response

REVIEWER, Line42: why speaking of ‘years after 2000’, so that in the 80^th and 90^th, there was already RAFOS floats and rotor current-meters on mooring lines to measure currents? In this sentence, a distinction could be made between Eulerian and Lagrangian techniques for currents measurements. Each have advantages and inconvenient.

AUTHORS: Agreed. A dedicated paragraph has been added dealing with Eulerian and Lagrangian techniques for currents measurements, in Section 1  (Introduction, starting from row no.65, pag.1 ).

*************************************************************************************

REVIEWER, Figure 1: the location of the buoy is difficult to see on the map.

AUTHORS: Figure 1 has been updated and the location is now easier to see.

*************************************************************************************

REVIEWER, Line 135: how have you found the uncertainties of 1.5 and 1.7 cm/s? 6.7 * 0.01 = 0.067 cm/s

AUTHORS: these uncertainty values have been calculated as the standard deviation of the dispersion of measures in field, in nominally identical experimental conditions for both the ADCPs : therefore, these values represent the repeatability contributions to the overall uncertainty during this ad hoc comparison.

*************************************************************************************

REVIEWER, Line161: how have you found the uncertainty of 0.20 dbar? Nortek pressure sensors are given to have an accuracy of 0.25 % of the full scale. The ADCP is a 200 m. Therefore, 200 * 0.25% = 0.5 dbar or 0.25 dbar if considering the specification as an expanded uncertainty.

AUTHORS: as declared, 0.20 dbar is not the standard uncertainty of the mean pressure measured by the ADCP, but just the standard deviation of the dispersion of pressure values along the whole deployment period. 

As reported in the manuscript, the pressure measures on board the ADCP are declared with a tolerance of ± 0.5 % of the full scale (200 m): i.e. ± 1 m. This value can be considered as a range in which the measurand (pressure) is uniformly distributed: therefore, under the hypothesis of a uniform (rectangular) ditribution, the associated standard uncertainty declared by the manufacturer is calculated as 1/sqrt(3) m.

The final standard uncertainty of the mean depth is then obtained by composing under square root both the contributions: RIPRODUCIBILITY: 0.20 dbar (or meter, it's almost the same) and ACCURACY: 1/sqrt(3) m...the final result is 0.6 m, as reported in the manuscript.

*************************************************************************************

REVIEWER, Lines 167-169 are repeated two times.

AUTHORS: ok, repeated lines have been removed.

*************************************************************************************

REVIEWER, Lines 252-253: could the 14 ° error be due to the positioning error of the ADCP by the diver, because it seems that the average error of figure 10 is close to 0 °?

AUTHORS: We added this missing information in the text in Section 2: "The frame of the ADCP was clamped by divers to pre-existing bolts. Such bolts have been fixed to the damping disk when the buoy was in the dry dock. In this way, it was assured that any frame attached to those bolts was perfectly aligned to the bow of the buoy."

Thus, the error of 14° cannot be ascribed to any misalignment due to the deployment by divers. Furthermore, in Fig.10 you can notice that the curve is not so symmetrical: the number of points in the first half (0-180°) is greater than that in the second half (180°-360°), thus justifying the average positive error of 14°.

*************************************************************************************

REVIEWER, Line 253: the battery packs can have also a strong influence on the magnetic field.

AUTHORS: as indicated by the manufacturer (see ref. [7], ADCP manual) a compass calibration was performed before deployment, to correct for the introduction of new magnetic materials (as the battery packs). The procedure, described in detail in paragraph 2.2.1 of the manual, was applied. 

*************************************************************************************

REVIEWER, Line 273: how is made the heading correction, point by point with the errors displayed on figure 10, or by using the average error?

AUTHORS: The heading correction is made point by point as specified in lines 257-258 of the original manuscript (the part just before Fig.9). To avoid possible misunderstandings, that info has been added also in caption of Figure 10.

*************************************************************************************

REVIEWER, Table 3: could you remember in the legend the significations of letters a, bc and O?

AUTHORS: ok, done in Table3 caption: “…Letter a stands for current measurements collected in 2004, while bc indicates the two time series merged for year 2005”.

*************************************************************************************

REVIEWER, Line 353-354: it must be an error in the reference numbers. Won’t it be [16] and [17]?

AUTHORS: Correct. References have been updated accordingly.

*************************************************************************************

REVIEWER, Line 365: how are calculated these standard deviations?

AUTHORS: the following part was added: "calculated as standard deviation of measure dispersions in the considered period".

*************************************************************************************

REVIEWER, Lines 388 and 391: the reference numbers are erroneous.

AUTHORS: Correct. References have been updated accordingly.

*************************************************************************************