I found the manuscript interesting and generally well written. In particular, the Introduction provides a clear and informative overview of the scientific background and motivation of the study. I recommend a number of clarifications and minor revisions that would improve the manuscript's clarity, reproducibility, and overall presentation prior to publication.

MAJOR COMMENTS:

- Please include a clear definition of the reported losses, preferably inlcuding the corresponding equation. The reference used to compute the percentage losses should be explicitly stated. Some related definitions appear later in the manuscript (Line 375), but this concept should be introduced earlier and described in greater detail.

- A map showing the geographical distribution of the considered sites would help the reader better understand the study area and the analysis presented in Section 2.

- Methodology: a description of the broadband radiation measurements is missing. Please provide information on the measurement sites, instruments/sensors, measured variables, quality-control procedures, and any other relevant details.

- How is GTI obtained from SSR? Please clarify the procedure used for this conversion, as this step can introduce significant uncertainties. If this calculation is performed internally within GSEE, please state so and briefly describe the methodology in the Section 2.

-The description of the GSEE implementation is not sufficiently detailed. Please specify explicitly which irradiance components (GHI, DHI, DNI and/or plane-of-array irradiance) were provided as inputs to GSEE for each simulation setup (measurements, libRadtran and METAL-WRF), and clarify whether the transposition to the PV plane was performed internally by GSEE. Also include some details in the PV generation model.

-  Figure 4: please clarify what is meant by "solar energy loss". Does it refer to tilted irradiance, PV energy production, or another quantity? Also indicate the reference used for the normalization and percentage calculations.

- Please include the optimum fixed tilt angle used at each site.

- Please provide more details on the clear-sky selection procedure.

MINOR COMMENTS:

- L35 - Please clarify that the sensitivity analysis is performed under clear skies conditions

- Fig.1. the figure is difficult to interpret. Is the color coding important for the analysis? If so, please clarify its meaning and improve readability.

- There is a missmatch between figure numbering and the figure references in the text. (e.g. Figures 5, 6 an 7). Please check figure numbering throughout the manuscript. Also, use a consistent format when referring to figures. For example Fig. X

- Ssec 3.3 Please refer to figures in order of appearance. For example, Fig. 11 appears referenced before Fig. 9.

- L260: fixed tilt: which angle?

- Fig. 9 and Fig. 11: please indicate which days are considered clear-sky cases.

- In the sensitivity analysis, please specify the wavelength associated with the reported AOD values.

- Fig. 11 and Fig. 1comparing outputs from a clear sky radiative transfer model (libRadtran in this case) with measurements affected by cloudiness is difficult to interpret and does not provide a meaningful assessment of model performance.

- conclusions: consider removing "Under cloudy conditions, the comparisons lead to greater differences and an overall overestimation in the case of libRadtran", as this result is expected when comparing clear-sky simulations with cloudy-sky observations.

SUGGESTIONS:

- Consider including a glossary of the main acronyms and variables used throughout the manuscript.

- L120: is there any estimate of the annual energy losses associated with dust events over the study region? Although outside the main scope of this work, such information could help place the reported event-scale impacts into a broader context.

- Since completely clear days are relatively scarce, the authors could consider using clear-sky periods selected from different days (even when a full day is not entirely cloudless) through the application of a clear-sky detection algorithm. This could potentially increase the amount of data available for the final analysis.

The authors presented a systematic analysis of the influences of dust aerosols on solar energy production in the Mediterranean Basin during four dust storms between 2021 and 2022. The topic is relevant, and the manuscript presents a comprehensive analysis. I would suggest a major revision before publishing on AMT. Please see my comments below.

Main comments:

1. The analysis is based on four different dust events. However, using only the station names listed in Table 1 makes it difficult for readers to understand the spatial extent and characteristics of these events. It is recommended to provide additional information describing their spatial distribution and intensity. Furthermore, as the manuscript indicates that solar energy production is strongly influenced by dust aerosol optical depth, it would be helpful to include a brief description of the intensity of each dust event.

2. The introduction section lacks a clear and coherent logical structure. For example, after the discussion of atmospheric aerosols, with dust particles as the main focus (lines 57-60), the authors return to a general discussion of aerosols' direct and indirect radiative effects (lines 60-61).

3. The authors estimate PV output using ground-based measurements and METAL-WRF for AMX and Lampedusa. However, given the spatial and temporal variation of dust transport and aerosol properties, it would be helpful to explain whether the conclusions drawn from these two case studies can be generalized to the broader Mediterranean Basin. Furthermore, as these two stations are affected by Saharan and Middle Eastern dust events, respectively, it would be useful to discuss whether dust from these distinct source regions has distinct implications for PV energy production and model performance.

Minor comments:

1. Lines 64, 75: Please avoid having two sets of parentheses next to each other.

2. Line 158: The reason for using Level 1.5 AERONET data should be provided.

3. Line 165: Need references for the criteria.

4. Line 365: It is not clear how the “global direct radiation losses” are defined.

5. Line 434: Why are only two stations used?

6. Line 515: “These days have been excluded.” Please be specific.

7. Please use either "Figure X" or "Fig. X" consistently throughout the manuscript when referring to figures.