This weird behavior occurred at 6 hPa (-60°C) and at 8 hPa (-70°C). Both times after decreasing the temperature and starting the new measurement campaign for different pressures.  The cause could be e.g. a time lag in the adaptation of the polymer to the new conditions (desorption of molecules or a phase change from water to ice) or a time lag in the adaption of the electronics inside the SHT85.

The paper can be expanded with these sentences to indicate the possible causes.

In the future, measurements at temperatures less than -50°C have to be scrutinized for this behavior  with a view to finding the causes.

Thank you for reviewing the manuscript.

1. “Please check on the numbering of the figures in the text. I guess in line 70 it should be figure 7 not figure 6.”

Yes, that’s true it should be figure 7.

1. “Equations 2 and 3: What temperature is T1?”

Yes, that was not mentioned. T₁ is the triple point temperature of water with = 273.16 K.

1. “Please add a short explanation why the measurements at -30deg have only been done at 8hPa.”

The measurement at -30 °C and 8 hPa was for orientation only. Due to the construction of the gas mixing and humidification system the highest reachable humidity is approximately 10°C to 11°C dew point at 1 atm pressure. This leads to a relative humidity of approximately 30% in maximum at -30°C and 8 hPa, which was deemed insufficient for a complete fit over the full relative humidity range. For this reason the measurements were not repeated at 6 and 10 hPa.

We will integrate our answers to point 1 to 3 in the manuscript.