Dear authors,
Thank you for taking most of the reviewer comments into account. However, some minor revisions and clarifications are still needed.

Especially the discussion of Fig 4 needs to be generalized. As also brought up by the first reviewer, this figure and its discussion only deals with one example of two ECCs, tested at three different weeks. However, in lines 380-381, you wrote that “a number of time-separated calibrations were conducted”. So, how many of these experiments have been conducted? And why do you not present the means of those experiments (i.e. average values at 0, 5, 10, 15, 20, 25, 30 mPa of those experiments) at the different weeks, as in Fig. 3, instead of showing just one example? Is the shown example representative for the other experiments as well? Please comment. Furthermore, the discussion focuses on the calibration for 30 mPa, which is, an unrealistic high ozone amount for the stratosphere. Can you also be conclusive for the finding that "the 30 mPa response of the ECCs increases with the week (compared to the reference)" for other ozone partial pressures? Please be more specific and more general.

In this context, your suggestion in lines 404-405 that “On the other hand, the changes could simply be a normal evolution of typical performance behavior” only holds if the illustrated performance is consistent (i) for the different time-separated calibration tested ozonesondes and (ii) for different ozone partial pressure levels (not only for 30 mPa). The same argument holds for the conclusion of this test in the summary section, lines 491-493 (“Results from testing ECC cells over a period of three weeks, one test each week, showed the calibration changed, e.g. about 10 percent for 1.0 percent KI and about 4-5 percent for the 0.5 percent solutions.”): it is not clear at all if this conclusion is valid only for the shown example and the numbers are true for all ozone partial pressure levels!

Line 44: already mention in the abstract that this study only deals with Science Pump Corp. ECCs

Line 72: write out EnSci here.

Lines 103-104: As asked by the second reviewer: use a more up-to-date reference for the accuracy of ozonesonde measurements here. E.g. the WMO GAW Report 201 (Smit & ASOPOS Panel 2014) gives an estimate (perhaps also Deshler et al., 2017; Thompson et al., BAMS, 2019).

Lines 119-130: provide the years at which the digital calibration benches were/are operationally used at Wallops, Payerne and Nairobi

Line 139: Write out TEI here.

Lines 240-243: please check the order of filling the cells. If at Wallops, like you wrote it down, the anode cells are filled before the cathode cells, the SOPs are not followed, which is a major issue for the conclusions reached.

Lines 288-289: “Generally, the downward calibration experiences small differences from the upward calibration”. In which sense? Consistent for all measurements? Related to the sensor response and consequently its memory of higher vs. lower ozone amounts?

Lines 325-327: “The final background currents often were somewhat higher than background currents experienced with manual preparation, generally about 0.04 microns” Please specify which final background current you mean here (final background current obtained prior to balloon release??) and give a possible reason why these background currents are higher with the automated procedure than with the manual procedure. Is this because of the linear calibration step in the automated procedure?

Section 3.2: you should mention in the beginning of this section that the 1.0% 1.0 B solution strength is the recommended one for Science Pump (e.g. Smit & ASOPOS, WMO GAW report 201, Deshler et al., 2017).

Lines 365-367: Please repeat here the explanation given by Johnson et al. (2002) about the effect of different KI solution concentrations and the side effects from the buffers used, within the context of your 1.0%1.0B & 0.5%0.5B SPC comparison.

Line 436: The profiles were averageD

Lines 436-443: what about the total ozone normalization factors of the 12 dual flights? Which solution strengths are closer to the co-located Brewer/Dobson or satellite overpass total ozone measurements?

Lines 456-460: what about the total ozone normalization factors of the three dual flights? Which solution strengths are closer to the co-located Brewer/Dobson or satellite overpass total ozone measurements?

Fig 04: I guess the units in these figures should be ranging from 0 to 30 mPa instead of 0 to 300?

Fig 06: I guess the legend should show the 0.3% KI in red, instead of 0.5% KI. As in Fig. 3, you might also list the standard deviations in this figure here. Is the smaller sample the only reason for the much larger standard deviations compared to Fig. 3?

Dear authors,

please take into account the following technical corrections:
* l 139: analyzer instead of analyzeer
* l 141: illustrateS
* l 253-254: you mentioned in your response to me why the downward calibrations are always higher than the upward calibrations. Please include this argument also in the text.
* l 381: Dobson measurements at Wallops, I assume? Please specify and you might also give a reference to a paper where the Dobson dataset at Wallops is presented.
* l 384: 20.9 DU instead of 20./9 DU
* Caption Fig. 5: 0.3% 0.3B instead of 0.3% 0.5B
* Fig 5: change the labels "1.0% mPa" and "0.3% mPa" for the blue and red curves (use % instead of mPa)!