The manuscript “Comparison of temperature dependent calibration methods of an instrument to measure OH and HO2 radicals using laser-induced fluorescence spectroscopy” by Winiberg et al. describes experiments to determine the temperature dependence of FAGE calibration factors. Two different methods have been used: (i) in an external calibration using a wand with a thermostated FAGE inlet with either room temperature or thermostated calibration gas and (ii) a chemical method taking place in the temperature controlled HIRAC chamber: hydrocarbon decay for OH calibration and measurement of HO2 decays due to self-reaction for HO2 calibration. The work has been carried out carefully and the results are presented in a clear manner. Even though the dependence of the calibration factor on temperature depends on the design of the FAGE system, and thus the obtained results are not transferable to other FAGE systems, the work is interesting to the FAGE community and I recommend publication after taking in account some minor comments.

Page 2, line 30 : an OH interference due to ROOOH has been identified on one FAGE instrument. This possible interference should be mentionned (Fittschen et al. ACP 19, 349-362 (2019))

p5,l22:typo

p6,l8 : what means quantitatively  negligible ?

p8,l16: is the H2O2 used  stabilized ? The stabilizer can be a source of interference in the FAGE. Did you observe such phenomenon  or H2O2 unstabilized has been used ?

p10,l10-12 and Fig 2 : description of the different conditions not clear; Fig2 (c) it says: the temperature is plotted when gas is either sampled from temperature-controlled air from the calibration flowtube or sampling from the HIRAC chamber. However, there is only one symbol for each OH and HO2. Slopes should be provided.

p12,l26 : sum of non-OH first order processes: provide examples

Fig 3a : could you add measurement uncertainties ?

p14, l10: could you comment the results obtained without mixing fans?

p14, l27: What is the order of magnitude of the percentage of HO2 typically lost to the walls rather than through self-reaction?

p16, l2 : be

p14,l4 : relative to the value at 293 K (to add)

p17, figure 5: I cannot see that C_HO2,obs data are more scattered. It looks like there is an increase at the beginning and then a plateau. What is the red line in the HO2 plot? Useful ? Wouldn't it be interesting to show uncertainty only due to temperature variation in addition to the total uncertainty?

p18,l22 : data points. Why only 3? Why are there more OH points than HO2 points?

p20: is the concentration of CH2O low enough to avoid HO2 loss through the equilibrium reaction HO2 + CH2O?

p20,l1:importance

Table S5 : explain k'rel

This paper presents the results of measurements of the temperature dependence of different techniques for the calibration of a laser-induced fluorescence – fluorescence assay by gas expansion (LIF-FAGE) instrument for measurements of OH and HO2. The authors compare the calibration factors obtained by the traditional water-vapor photolysis technique as a function of temperature with that obtained by hydrocarbon decay by reaction with OH for OH calibration, and measurement of HO2 decays due to self-reaction for HO2 calibration. The authors find that all calibration methods had a positive temperature dependence. After accounting for known temperature dependent factors, such as number density, quenching and rotational population of the probed level, the authors conclude that the remaining temperature dependence is most likely due to a decrease in the wall loss of radicals in their instrument as the temperature increases, improving the transmission of radicals from the pinhole inlet to the detection axis.

As noted by the authors, these results depends on the design and materials of the FAGE instrument and may not be applicable to other FAGE instruments. However, the results are of interest to the HOx measurement community and suitable for publication in AMT after the authors have considered the following minor comments.

• As mentioned in the manuscript, the calibration factor will have a dependence on water vapor due to quenching of the fluorescence. The authors illustrate the OH calibration factor at a constant concentration of water, but there is no mention of how the authors account for the water vapor dependence of the calibration factor. The authors state that the HO2 experiments were done in dry conditions to minimize enhancement of the HO2 self-reaction by water vapor. I assume that the water vapor calibration factor was corrected to account for quenching by water vapor when compared to the calibration factors derived by the other techniques, but the authors should clarify how this was done. Did the authors measure the calibration factor as a function of water vapor in order to correct the calibration for dry conditions?
  
  
  
  
• There appears to be an error in Figure 5b, as the points do not reflect the values in Table 2, although the fit does.
  
  
  
  
• The authors should clarify how the fits of the temperature dependence shown in Figures 5-7 were obtained. Are these bivariate fits weighted by the measurement uncertainty?
  
  
  
  
• The errors in Table 4 should be clarified – are these 1 sigma uncertainties?