General comments:

The authors have provided a comprehensive experimental study of the reactions of sabinene with OH and O₃.  The novel contributions include measuring the temperature dependence of the OH reaction and analyzing the chemical budget for OH radicals.  The authors take pains to present previous work, describe their experimental and data analysis methods, and compare their results with past experimental and theoretical values.  Based on the scientific significance and quality, I recommend accepting the manuscript.

Specific comments:

The authors may want to consider other ozonolysis reaction pathways that may affect OH yield.  This includes decomposition of chemically activated CH₂OO that produces OH (see Pfeifle et al., J. Chem. Phys. 2018, 148, 174306) and rearrangments of vinyl hydroperoxides that may reduce OH yield (see Barber et al., J. Am. Chem. Soc. 2018, 140, 10866-10880).

The authors mention a signficant discrepancy between their measured sabinene + O₃ rate constant and the SAR value from Jenkin.  Some additional discussion of the possible origin of this discrepancy would be illuminating.

Technical corrections:

It would be helpful if the authors briefly summarized the theoretical methods employed by Wang and Wang.  This would give the reader some sense of how reliable the theoretical predictions are without having to look up the references.

There may be a few too many details presented in Section 3.3.

The LIF instrument, since its 1σ uncertainty is smaller than that of the DOAS instrument, has the higher precision (contrary to line 207).  Along similar lines, the relevant column heading in Table 1 is better termed uncertainty than precision.

There are a few typographical errors to correct.

The rate coefficient over a range of temperatures was measured in the laboratory for OH + sabinene using a field OH reactivity instrument, and an Arrhenius expression obtained. Also, an impressive suite of state-of-the-art instrumentation was used in the SAPHIR chamber at ambient temperature (including down to 278 K in winter) to measure concentrations of trace gases (reactants, products, intermediates and oxidants, including radicals) following the oxidation of sabinene by ozone (dark experiments) and OH (photoxidation experiments) at various levels of NOx in the SAPHIR chamber. There were also accompanying measurements of physical parameters (e.g. T and J values). OH was measured using two methods (DOAS and LIF) and HO2, RO2 using chemical conversion without/with an external reactor. As well as rate coefficients, yields of products (HCHO, acetone, sabinoketone), including OH from ozonolysis experiments, were performed. An OH budget analysis is also performed in which the rates of destruction (via OH reactivity and [OH]) and production were experimentally measured – allowing examination of whether missing OH sources are present – or whether rate coefficients for production processes already included are correct.

Many experiments were performed, and the dataset obtained is comprehensive and impressive. At ambient temperature the rate coefficient for OH + sabinene between the lab and the SAPHIR chamber was in good agreement. Having a rate coefficient measurement both from the lab and a chamber in the same study is to be commended, and shows the strength of multiple approaches. The sabinene + O3 rate coefficient at 278K was also determined in the chamber in winter experiments, and was less than a previous determined value at room temperature. The SABINOHBO2 RO2 species is unusual in that the unimolecular isomerisation rate is fairly large, meaning that even in polluted conditions this is a major loss route for this RO2 radical. However in this work the isomerisation rate obtained experimentally from the chamber experiments is not as large as from the previous calculated ab initio value.

A careful comparison with previous work is given, in many cases there is good agreement (e.g. for rate coefficients and yields), but not always, but the reasons for the differences are discussed critically. In this paper there is no comparison with any model calculation, rather rates of production are calculated from the measured data, and the steady-state approximation for OH is used to infer OH production rates (not measured) from measured OH loss rates (OH reactivity x [OH]) and some measured OH production rates. The data are of good quality, and are discussed fully within the existing literature (e.g. previously developed mechanisms). This value was obtained by using the measured yield of acetone as a target and adjusting the isomerisation rate to bring the calculated value (from a budget analysis) into agreement with the measured value.

The paper is very well written and very clear to follow. The paper is suitable for publication in ACP. The authors should consider the below in any revised MS.

General.

• Rate coefficient preferred over rate constant, although the latter is used in common parlance.
• Bimolecular rather than Bi-Molecular (and elsewhere) , you have unimolecular already without a hyphen
• Are there plans to use a model to calculate the radical levels, or products? Will a mechanism for sabinene oxidation be developed to extend for example the MCM?

Abstract.

• Line (L) 13. exp((575±30)T-1) would be better as exp((575±30)/T), also on L390, and elsewhere (e.g. Table 4)
• ... the analysis of (not if)

Introduction.

• O3 (Wang and Wang, 2017), ..... add respectively

Methods

• temperatures rather than temperature
• In situ and not in-situ
• L128, it is probably worth adding followed by reaction of the O(1D) atoms generated with water vapour present in the gas mixture
• L138 Sabinene is measured using a TOC instrument with detection of the CO2 formed after pyrolysis. Is there good evidence that 100% of the sabinene is removed to form CO2? Also, is the reason why a TOC is used for the high time-resolution required? It may be that the time-resolution is not good enough for the experiments for other methods (e.g. MS) – but perhaps a note to note this is needed.
• L144 and L148. Temperatures here are in Celsius, whereas in the abstract they are given in K. I suggest that the abstract retains K as given, but perhaps K in brackets after the T in Celsius might be useful to add. Otherwise it will be confusing as for equation (2) on line 158 the T here has to be in K.
• Please state the zero reactivity of the OH reactivity instrument.
• Replace -EAR-1 by -E_A/R to be consistent with the equation
• Please provide a reference for the ROxLIF instrument.
• L200-201. Is the HOx cell the same as the second LIF cell where HO2 is measured? Is the ROx cell the same as “ another low-pressure LIF detection cell”. Make this clear if the case. as HOx and ROx cells are not defined at present
• L202-207. It is excellent that there are two independent in situ methods for measuring OH. This is unique to SAPHIR. However, the statement that the DOAS and LIF instruments agreed with each other is a bit confusing if the difference between the 2 instrument is similar to the concentration of OH during the experiments? This suggests a significant difference? I agree though that because the OH concentration is similar to or below the 1-sigma precision of the DOAS instrument, that actually only the LIF instrument value is used.
• L212-213. DOAS is used for the OH measurements in the photooxidation experiments, and there is a note that the difference with LIF might be due to an unaccounted calibration error. Is there confidence then in the LIF value for the O3 experiments when only the LIF value is used?
• Can a note be added about why the PTR was not calibrated for sabinene. Is there an experimental limitation which prevents this?
• L228, factors not factorss
• There are quite a few Novelli et al., 2023 references (a, b, c, ....), and specific figures are referred to in these (which are Eurochamp datasets), e.g. from Table 2 and also in the text. The figures were all blank for me when I clicked on the links (the axes show but no line or point on the graphs). I wonder if the Supplementary Information (SI) for this paper could be used to display these specific figures instead of having to click on links (which did not any data for me) – it would be easier for the reader? I think if the intention of the references is to contain significant supporting data for each experiment, then this is fine (but the figures were blank for me), but if specific Figures from these databases are cited, then these ought to be more readily observable via the SI.
• Figure 3. For the last panel, the loss of RO2 seems to be completely dominated by RO2+HO2, with RO2+RO2 being very small indeed. Given that the concentrations of RO2 and HO2 are similar (two other panels), then I think a comment is needed to say that the RO2+RO2 rate coefficient is 20 times smaller than for RO2+HO2 (this information is in the SI).
• Add (100 ppmv) after CO to make this clearer again.
• The MCM is used to calculate photoylsis rates of ketones – can a reference please be given for this.
•  

Results and Discussion

• Known and not know
• “OH production rate was excellently balanced”, better might be “.... “ the OH production was very well balanced...”