Review of Gramme et al. (2025): "Urban pollution monitoring with the AOTF-based camera: NO2 validation with other DOAS instruments"

The manuscript presents a novel ground-based NO2 imaging instrument based on acousto-optic tunable filter (AOTF). The new instrument builds on a previous AOTF-based instrument and aims at capturing high spatial and temporal variability of the reactive NO2, a key trace gas for air pollution studies. The capabilities of the new instrument are described and tested during a field campaign in an urban environment in Italy. The imager was mounted next to two DOAS-type well-established instruments (MAX-DOAS and Pandora). NO2 differential slant column densities (dSCD) obtained after performing the DOAS analysis of the three instruments’ spectra are compared and their differences assessed. Overall, the imaging results show good agreement with the reference instruments (strong correlation for SCD observations), demonstrating the potential of the new instrument in the context of urban air quality monitoring. The authors assess the limitations of the new instrument and suggest future work to keep on improving the technique and its applications.

The manuscript is well presented and the methodology is well described. Figures are clear and support conclusions. Overall, publication is recommended after addressing the following comments in the manuscript to further strengthen its impact.

General Comments:

The manuscript could benefit from a more detailed comparison with other NO2 remote sensing instrument, particularly alternative imaging approaches such as grating-based imaging DOAS, gas correlation cameras (e.g. volcanoes studies) or Fabry-Perot systems. This would help to frame the novelty of the instrument presented.

The manuscript uses NO2 dSCD for instrument comparison and, towards the end of the draft, the authors state a future use of the O4 bands for light path assessment. However, it could help the reader if the authors could clearly state at the beginning of the manuscript that the relationship between NO2 dSCD and actual concentrations of NO2 depends on several factors (e.g., aerosol, clouds, surface albedo), which may not have the same effect on all the three instruments. Indeed, including O4 band and inversion of NO2 concentration would be crucial for a thorough assessment of the capabilities of the (three) instruments, their detection limits and their comparability.

Also, the authors compare the new instrument with the two reference instruments and later on, they compare the reference instruments between themselves (last paragraph of Sect. 5.2). However, the authors are comparing three instruments, with three different field of view and two different DOAS (reference type) analysis. Also, as shown in Table 3, three different DOAS settings are used for the analysis. All these differences will affect the comparability of the data. The manuscript and conclusions could benefit if the two reference instruments were firstly compared with each other so a baseline for the quality of the comparison could be set. This could e.g. state (and quantify) the limitation of comparing NO2 dSCD analyzed with a fixed zenith reference spectrum (new instrument and MAXDOAS) with those analyzed with a sequential reference spectrum (Pandora).

Specific Comments:

P3, L52: “…, many more wavelengths are now acquired in routine operations in order to achieve higher accuracy”. Please be more specific.

P3, L56: Instrument description. Is the instrument calibrated radiometrically? How does temperature changes affect the AOTF spectral tunning?

P3, L70: “more compact”. Please provide size.

P6, L118: “In that case, a 1-D NO2 field can be observed”. Do the authors mean “can be observed per azimuth?”. In that, case, would it be a 2-D field?

P6, L124: “…capture complete images of the scene, but one wavelength at a time”. How long does it take to gather one image at a given λ before moving to the next λ?

P7, L128: “lower signal-to-noise ratio”. Does the instrument have a SZA limit where the SNR is far too low to assure the quality of the observation? Given the case, please provide such SZA limit.

P7, L131: “… focusing on a limited number of wavelengths”. How do authors decide how many λ are a minimum? Please specify the rationale behind.

P8, L161: “PRNU” How is the PRNU characterized?

P9, L166: “cosmetic step”. What do the authors mean with that? Do you mean the PRNU correction could be omitted? Is PRNU 100% independent of light intensity?

P14, L266: “lowest elevation angle”. Which angle? Elevation angle of the Sun or of the telescope? Clarify

P14, L267-269: The way it is written how the MAX-DOAS is compared to the Pandora observations seems a bit confusing. Please, detail the procedure in more detail.

P16, L295-296: “not shown”. Including the comparison with Pandora could assist the reader and the conclusions (supplementary?).

P16, L296-297: “…short slant column of NO2 between them and the instruments.” What do the authors mean? A shorter light path?

P17, L325: The authors refer to the effect of “moving clouds”. How would aerosols affect?

P21, L346: “Given the unconventional concept of the NO2 camera...”. What do the authors mean by “unconventional”?

P21, L361: What is the result of those changes in illumination conditions?

P22, L373: “pollution events” are often linked not only to NO2 but also to particles. The draft could benefit if the authors could assess the possible effect of aerosol on observations.

Technical Corrections:

P1, L2-5: “Existing instruments….. hot spots and quick variations”. This is a rather long sentence. Consider splitting it.

P3, L53: “In the following section,…”. Please change to “In Sect. 2.4,…”

P3, L70: “RF”. Define acronym.

P5, L97: “Fig. 6”. In the text of the manuscript, “Fig. 6” appears before Fig. 3, 4 and 5. It would ease things for the reader if the number of the figures were re-numbered.

P8, Fig. 5: Please add the units of the vertical and horizontal axes

P11, L214: “PTC”. Define acronym.

P14, L262: “In order to dSCD comparable…”. Consider rephrasing it “In order to compare dSCD from…”

P14, L272: “The each different…”. Change to “Each different…”

P17, L327: “comparison made earlier”. In Table 4?

P20, L331: “Sections 3.4, 4.2.2 and 4.2.1”. Better change to “Previous sections”

P20, Table 5: Please, specify for which elevation angle are those values

Gramme et al. present a new ground based NO₂ instrument, which in contrast to conventional differential optical absorption spectroscopy (DOAS) applications does not measure the whole spectrum at once and instead measures a whole 2D image at a specific wavelength and shifts through the wavelengths using an acousto-optic tuneable filter. Each image represents the measured intensities at a specific wavelength, which in combination can be evaluated using the DOAS technique. This allows the instrument to capture a 2D NO₂ field every few minutes. The instrument is based on a proof-of-concept optical breadboard of the VIS channel for the ALTIUS mission and has been refined and improved to work as a ground based NO₂ instrument. The instrument was set up next to and compared to two state of the art ground based remote sensing instruments, a Pandora and a MAX-DOAS. NO₂ differential slant column densities are obtained and compared to each other for each instrument. The NO₂ camera shows a good agreement with the two reference instruments. The capabilities of the NO₂ camera are highlighted and possible issues are discussed. Generally, the manuscript is well written and the methodology is well explained and I will recommend it for publication, but beforehand I would suggest some minor improvements to the manuscript.

General comments:

If I understand correctly the camera needs a little longer than 188 seconds to measure all the data cubes necessary for a single dSCD. How are the dSCDs affected if there are highly inhomogeneous illumination conditions, e.g. when there are fast moving clouds all over the horizon?

Table 3 shows the different fit settings for the different instruments, which are slightly different for each of them. I assume there is a good reason for the different fit windows, polynomials etc. for each instrument, but maybe you can elaborate a little bit on this. Is there a specific reason for using different O2 dimer cross-sections?

In the estimation of the uncertainty of each NO₂ dSCD, the systematic contributions are ignored so far, but maybe you are able to provide a rough estimate for some systematic contributions already.

Specific comments:

L266: I am little bit confused by this, does the PGN use a low telescope elevation angle measurement as the reference? Please clarify this part.

Technical corrections:

L53 – L 55: These two sentences are similar and I guess one of those is leftover from a previous iteration of the manuscript.

Table 1: Table caption should be above the table.

Table 2: Table caption should be above the table.

L123: “Instead of scanning the scene, (Dekemper et al., 2016) proposed…” remove the () around the citation.

Table 3: Table caption should be above the table.

L219: I assume u_z,L1 and I_z,L1 are the uncertainty and Intensity of the zenith measurements, however I think it’s good practise explain all variables in an equation.

L259: “… its outputs is provided in (Cede et al. 2025) and …” remove the () around the citation.

L262: The first part of the sentence is missing a verb.

Figure 8: Please check if this figure is suitable for red / green color blindness.

Table 4: Table caption should be above the table.

Figure 9: the titles of the subfigures are clipped

L331: “Sections 3.4, 4.2.2, and 4.2.1 above…” change the order of the sections so they appear in ascending order.

Table 5: The table caption should be above the table.