Review of “Correction of the Observatoire Haute Provence electrochemical concentration cell (ECC) ozonesonde 1991-2023 data record” by Okamoto et al.

Summary and General Comments:

The authors summarize ECC ozonesonde measurements collected at the OHP station beginning in 1991, including a period of anomalously low ECC pump temperatures from January 2001 to July 2007. In the absence of any other changes, erroneously low ECC pump temperatures will cause measured ozone values to be too low. However, the authors were unable to find an explanation for the low pump temperatures, and state that no changes to the ozonesonde preparation were made during January 2001 to July 2007 to cause them.

The ECC pump temperature adjustment and subsequent recalculation of the ozone values are performed using near-coincident OHP lidar stratospheric ozone as a reference. Using the OHP ozone lidar data to adjust the pump temperature and resulting ECC ozone measurements has the unfortunate consequence of losing the independence of the ECC ozone measurements during this period.

I am not aware of similar pump temperature behaviors at any other En-Sci ozonesonde station during this period, so I strongly encourage the authors to expand discussion that rules out potential causes and seek to find an explanation for this anomalous behavior. For example, are the ECC pump motor current and ECC battery voltage recorded for this time period? Was the pump thermistor instead taped to the side of the Styrofoam box instead of the pump block during this period? Also, it was discovered a few years ago that the ozone calculation scripts in Vaisala’s MW41 software had a bug that reported incorrect ECC pump temperatures, causing erroneous ECC ozone measurements. Could a software issue potentially explain this?

I also wonder about the validity of a linear correction to the pump temperature. Have the authors evaluated the corrected ECC ozone measurements other than in the stratosphere and total column? Also, it is not explained how the total column is derived from the ECCs – I presume an above-burst satellite climatology is used.

The authors should consider changing 1991-2023 to 2001-2007 in the title, since those are the only dates where the ECC measurements are adjusted.

Recommendation:

The paper is concise, well-written, and only requires a few minor technical corrections. However, before acceptance I strongly encourage the authors to expand the discussion on the potential origins of the anomalously low ECC pump temperatures from 2001-2007, and the validity of the linear pump temperature correction.

Line-by-Line and Technical Comments:

• Abstract Line 4: “instruments”
• Abstract Line 5 (and main text): Clarify that this is Aura MLS at first mention
• Abstract Line 9: Not really the “pump speed trend” but the pump flowrate distribution, correct?
• Line 62-63: “No significant sounding preparation changes occurred during the time period March 1997 and July 2007.” However, I must believe there is an explanation for this pump temperature behavior that is yet to be discovered
• Line 73: Change “when” to “where”
• Figures (general): Higher resolution figures should be submitted
• Line 98 and 128: Be precise in whether you mean “statistically significant” or not, and how that is defined
• Line 106: I think Figure 3 shows the opposite of what this text describes
• Lines 107-108: It is not clear to me why changing the radiosonde, interface, and batteries would cause the ECC pump temperature to change. Certainly, a change in the thermistor location would do that, if that is what you mean
• Figure 2 caption: There are only panels a and b in this Figure
• Figures 3 and 5: Please use the same color scale on both of these Figures. I prefer the Figure 5 colors
• Table 1: I assume these are values for 25km as also indicated on Figure 4, correct? Please make that clear in the Table 1 caption