The manuscript by Bevilacqua et al. is an interesting analysis of historical and science documents that aims at producing a quantitative dataset of where volcanic ballistic projectiles (VBPs) impacted during major explosions and paroxysms at Stromboli volcano. Starting from previous databases of past eruptive events, the authors define a few parameters that effectively measure the range, direction, and dispersal of VBPs, also providing an associated error. The results are discussed mainly in terms of the methodology and associated sources of uncertainty, in the perspective of a companion manuscript that will use these results to generate hazard maps for VBPs at Stromboli. It is very interesting to read about how the authors approached the difficulty of extracting useful information from such a diverse span of sources. The cursory discussions on the distinction between activity styles and the factors controlling VBPs ejection are valuable, although slightly simplistic.

Editorially, the manuscript is very well written and clearly illustrated.

To improve the manuscript, I suggest spending a few more words on the Monte Carlo simulations, how they are used to estimate uncertainty, what is the rationale for choosing them and the steps followed to implement them. It would also be beneficial to add some comments and references on the use of the VBPs range in dividing major explosions from paroxysms. This risks appearing to the reader as a circular reasoning, since the distinction is made a priori but then the VBPs ranges for the two styles are shown to have no clear divide. Finally, a few words could be used to emphasize how much bouncing and rolling contributed to the final range of VBPs on the steep slopes of Stromboli, a problem possibly unknown in some flatter volcanic areas.

Minor corrections and suggestions are as follows:

Line 8: analyzes >> analyses, I guess. Or do you mean Analyzes (Verb - US): Third-person singular present tense of the verb analyze?

Line 11: Our findings highlight that>> Our findings highlight that: (1). Since later you have a (2), you need a (1) here.

Line 28: Ballistic projectiles>> Volcanic Ballistic Projectiles. More accurate terminology.

Line 29: Missing stop mark after ‘activity’.

Line 254: 3.1 Analysis of maximum distance, direction and area affected by ballistic projectiles IN MAJOR EXPLOSIONS. This section deals only with major explosions.

Line 348: ‘and’ instead of ‘+’.

Line 423: ‘…it would have been impossible to...’: This sounds a bit too strong. Other simplified approaches could have worked equally well.

Line 457: …’ able to re-shape the shallow conduits and craters…’. For activity change related to paroxysm at Stromboli see Salvatore, V., Silleni, A., Corneli, D., Taddeucci, J., Palladino, D. M., Sottili, G., ... & Cristaldi, A. (2018). Parameterizing multi-vent activity at Stromboli volcano (Aeolian Islands, Italy). Bulletin of Volcanology, 80(7), 64.

This paper presents a novel mapping technique used to estimate the volcanic ballistic projectile areas for 67 past events at Stromboli. Overall, it is well written and the method, but more importantly the resulting dataset, are worthy of publication.

I have no major issues with any part of the paper, although it is slightly annoying that it is a part I of II - I presume part II will be the more exciting read. Mainly I'd just like a bit of clarification around some of the later figures.

Below are specific minor comments / typos:

Fig 1. It is quite hard to see the inhabited areas in (i) – a different colour might help, or just not include that information at all. There should also be some more emphasis placed on the fact that to the NW, you wouldn’t see any ballistics greater than about a km because they’d be in the sea, but you would see them in any other direction.

L91 – “deduced” might be better than “deducted”

L92 – “cameras reports” à “cameras’ reports”

L126 – I didn’t quite follow the meaning of this line (“distances and angles are not related to the vent ejecting the projectile but to a fixed centre”) à is this just that you used one central location and ignored the fact that the vent location varied?

L141 – “vents” is used here – so for each individual one you used the individual vent, but not when combining?

L160 – “and of major explosions” is missing a number or proportion here

L162 – suggest removing “precautionary”

Table 1 – “between0”

Figures 5, 6, and 7 are great, but I was then expecting a final one with them all overlayed on each other? Maybe one for all of 5, one for all of 6, one for all of 7, and then one for all of all? It would just help to have a qualitative “by-eye” type comparison to align with the plotted results later on.

Figures 8 & 9 I found quite hard to understand. They look really useful, but (for example), why are the uncertainty dashes only on one side of the dots, why don’t the green dots consistently have larger uncertainties than the blue dots. Where did these uncertainties actually come from? Why mark three paroxysms on the major explosions plot?

Figure 10 – shouldn’t the uncertainty be in the directionality (i.e., L-R) on (c) and (d), and not in the probability? These also uncertainties seem very small. I presume also that the result “more major explosions go to the SE” in (c) is not purely because if they went NW they’d fall in to the sea?

L455 – suggest removing this line “As a matter of fact, the areas affected by ballistic fallout of past explosions are essentially characterised by past data”