General comments:

This paper is written on the development, characterization, and application of a marine aerosol generator. The authors provide a very thorough review of relevant literature both in the introduction and throughout the manuscript. Multiple tests were conducted to validate the results of their system, while also comparing to results of similar systems. The system is then applied to several samples to compare their size distributions, chemistry, and INP concentrations. The results are thorough and make for a linear story as well. 

Specific comments:

• Lines 164-166: Cite or mention Stokes et al. as reference for system
• Line 178: Provide brief explanation on why the chosen cascade was the most suitable.
• Are OPC counts biased towards the lowest bins?
• 3.1: Which cascade was used for the background particle test?
• 3.1: At what (time) point does the SSA reach a steady state? This information would be useful.
• 3.2: Which cascade is the best/most realistic? Should be explicitly stated here.
• It should be mentioned how the longer delay also allows for resettling/reformation of the SML, which is important for the chemistry of the particles.
• Line 179 states that with 40 L of water, the waterfall is roughly 10 cm, however Lines 401-402 state that a 40 L fill results in a 22.5 cm waterfall height. This is a 2x difference. Please address this discrepancy.
• Line 464-465: Not necessary to state here that coarse mode particles can be produced. This point is made and should be stated in section 3.5
• The authors reference how similar works induced phytoplankton blooms and that is what accounts for varying INP concentrations in other works. A comparison of INP concentrations from their work to INP concentrations from the start of the blooms (prior to culture additions) for other works would be a more appropriate comparison and useful addition.
• Centrifugal pumps can be harsh on biology and this could be affecting overall INP concentrations. Have authors considered this and measured if the number of INP at specific temperature changes over time? Another option is to perform experiments with other pumps to test affect on INP concentration.

Technical comments:

• Line 74: Delete “in.”
• Line 120: Suggest using “species” instead of “particles” as most particles can contain more than just a diatom or dinoflagellate.
• Line 190: Water flow or air flow?
• Figure 1: This is meticulous, but only 8 screws are accounted for in diagram and text states that 10 are used to hold lid.
• Line 275: Section 2.4 label
• Line 334: *produced
• Line 481: Should be decreases, not increases
• Line 613: *tank

The paper describes a newly designed marine aerosol tank used to generate simulated sea spray aerosol in laboratory settings. The authors give a comprehensive overview of other aerosol tanks in the introduction and carefully describe their results in the context of other tank studies. The tank setup is described in detail, including a clear figure. The tank was tested using both artificial seawater and seawater gathered from the Gulf of Mexico. The analysis of experimental results is given, including both chemical and physical characterization of the aerosol generated from the waters. While results are described in the context of other aerosol reference tank studies, the results from this study are not described in clear detail. More quantification should be given, and the authors should be careful to avoid vague language such as stating that “a better correlation” was found rather than giving exact values. With revision, this paper represents an interesting addition to lab-based sea spray aerosol methods.

L403: “likely due to the decomposition of organic matter” – without a measurement of the organic content in water at the time of measurement, this cannot be stated definitively, as turbidity could also be due to suspension of inorganic particulate matter.

L443: it needs to be considered in the analysis that water samples sat at room temperature for fourteen days prior to analysis. This will have a major effect on water chemistry since only organisms >50 µm were filtered out. This would essentially remove all grazers and zooplankton from the system while leaving phytoplankton and bacteria to live on the collected nutrients for fourteen days and cannot be considered truly representative of natural water composition. Nitrate, ammonia, and other nutrients will be metabolized by phytoplankton as the water sits, artificially lowering the concentration.

L535: Missing a word between the two phytoplankton species

L541: It may also be that the samples used in this study contain a greater proportion of decomposed or dying material than the two studies listed. While this is addressed later in the paper, it should also be considered here.

L557: Please give actual correlation values and discuss statistical significance of given values

L565-582: This should be considered earlier in the paper as impacts are likely to impact all results given.