General comments

Marini et al. attempt to integrate new gas geochemical data and existing information into a revised “conceptual model” of the Solfatara magmatic-hydrothermal system at Campi Flegrei. I found the manuscript very difficult to review as a lot of important information, namely regarding the methods, the underlying assumptions, uncertainty assessment, known vs. inferred geology, as well as the exact lines of reasoning from data to interpretation are missing. This renders the work in its current form essentially irreproducible and very difficult to review because essentially almost every paragraph can be challenged. Even some (central) elements of the discussion seem to contradict each other; e.g., the concept of separate “reservoirs” as locations of equilibration of different gas geothermometers is obsolete with the assumption of flow through a vertically connected structure. Also, regarding the core chemistry, the authors (apparently coming from a volcanology background) seem to rely on shallow fumarole gas equilibria and ignore stablished knowledge from the magmatic-hydrothermal systems community for the deeper and higher pressured parts of the system (e.g., H2S generation through the well-known SO2 disproportionation reaction). I think that for these reasons the manuscript should not be published in its current from and requires a complete revise beyond “major revisions”.

In the following I will highlight some key problem areas (and for each one only some facets and ignore many questionable details) of the manuscript but will refrain from a detailed list of all issues.

Specific comments 1: methods and assumptions behind them

The authors rely on three gas equilibria to obtain “distinct equilibration temperatures and related total fluid pressures”. They do not elaborate how these were computed and not even how such a “equilibration temperature” is to be interpreted (e.g.: is it the lowest temperature at which the reaction went to equilibrium along the fluid’s upflow path and which was then preserved due to kinetic limitations at lower temperature?), and how such equilibrium could be demonstrated at all.

It remains completely opaque how the computations were done. I would argue that at least some of the stated computations are currently not meaningfully possible for the lack of adequate thermodynamic data and models, e.g. for a complex gas phase in equilibrium with a 21 or 33.5 wt% aqueous NaCl brine – for the conditions stated there exists neither an adequate solubility model for the gases in such a brine nor are even remotely accurate activity or fugacity coefficients available. How then could temperatures and pressures be derived? I tried to make my way through the main author’s 2022 book (375 pages, the main findings of which the present manuscript is supposed to summarize and explain) but I could not identify a coherent summary of the computations and the underlying assumptions with reasonable effort. This manuscript should deliver a crystal-clear and reproducible description and openly lay out and discuss the major assumptions to become a publishable contribution.

There exists a relatively elaborate assessment of sulfur chemical behavior in the deep vs. shallow parts of magmatic-hydrothermal systems that was established some decades ago in the economic geology community (just to name the works by Einaudi and Hedenquist as an example) including the importance of “rock-buffered” vs. “direct degassing” paths. The very relevant insights presented there go largely ignored by the present authors. Namely, the important SO2 disproportionation reaction to generate H2S (and there are variations of the theme depending on temperature, rock, etc.) is ignored although it is considered most important by many researchers in the field. The reactions (3) and (4) involving calcite also seem odd; most people working on the deeper parts of magmatic-hydrothermal systems would see these reactions going to the left (albeit possibly rather involving SO2 and using plagioclase rather than calcite) to sequester sulfur rather than creating H2S, see, e.g., Henley at al., JVGR 2022. I wonder what “H2S temperatures” would come out if this was properly taken into account of if “H2S temperatures” would have any meaning then.

Specific comments 2: the core of the conceptual model

When looking closer one has a hard time to understand reasoning behind the “conceptual model”. On the one hand, three “reservoirs” at different depth, separated by aquicludes, are postulated; on the other hand – to make the fluids migrate from a magma at >8km depth to the surface – connected flow through a vertically extensive “structure” needs to be invoked.   

Now, first and foremost, I had a hard time understanding why the different gas equilibration temperatures should be specifically connected to the individual “reservoirs” and how the latter were inferred in the first place. Is there some circular reasoning? Or does actual, drilled geology with porosity and permeability measurements come into play? There is some mentioning of wells etc. but the exact reasoning remained blurry to me, at best. So, one cannot even tell if the depth of “reservoirs” and “aquicludes” is well-constrained or just a guess.

Then, if there is a vertically connecting structure – what is the meaning of the temperatures then? “Equilibration” will happen (and be “frozen in” if my above speculation about the meaning of “equilibration temperature” is correct) somewhere along the flow path and why should that be connected to the depth of any of the “reservoirs” then? Wouldn’t it make more sense that the increase in apparent temperatures and pressures (if correct) reflects rather a change in the hydraulic regime (or the degassing rate) such that the chemical signal of deeper fluids gets better preserved rather than a specific “reservoir” getting hotter and stronger pressurized (see, e.g., the overpressure waves in a magmatic-hydrothermal systems self-developing in the simulations of Weis et al., 2012; later suggested also by Lupi et al. for Campi Flegrei). What would speak for and against such different possibilities, why aren’t they considered and also tested against the data?

In the whole conceptual model discussion, speculative ideas (such as Fournier’s “self-sealing” quartz layer for which later studies found little evidence), inferred vs. drilled geology, geochemical data with different possible interpretations etc. are just mingled without testing for plausibility etc. Personally, I think that this is quite far away from best practices; the different ideas should be formulated as hypotheses and then tested to the degree possible.

Specific comments 3: drawing straight lines without reasoning

The temperature and pressure profiles in Figs. 5 and 6 hinge on the – untested – assumption that the gas equilibration temperatures are representative for three “reservoirs” at different depth. In Fig. 5, temperature in each reservoir is taken to be constant. Unless I missed something important no reason is given why this should be the case. Rather, it is taken as granted (out of nothing) and then it is postulated that convection inside the reservoir homogenizes the temperature. Between the reservoirs – again: unless I overlooked something important – straight lines are drawn without reasoning and then it is stated that the “the heat transfer appears to be controlled by conduction”. I think this is quite poor scientific practice to just draw a straight line in the absence of data and then to assume it is correct and make such a conclusion.

For fluid pressure (Fig. 6) the “reservoirs” are also drawn to have constant pressure even if more than a km high. This is obviously unphysical as there would have to be a hydrostatic pressure gradient (not necessarily linear as density of the fluid may vary with depth).

Some other specific comments

I only list a few, which I think are important:

• 1: I think that a Google Earth snapshot is a no-go and contains much less info than, e.g., a simplified line art map showing the main geologic structures and the main geographic locations. Add a square that show the location of Fig 1b. In 1b show a scale rather than an unexplained coordinate system.
• Avoid self-celebrating statements in the introduction.
• “standard deviation” in line 98 includes what? Just the effect of variable gas analysis? What’s the uncertainty of the thermodynamic analysis to obtain T and P? Possibly much larger?
• I can’t follow the reasoning in 4.3; if you assume that (1) and (2) work well separately, than (5) should work as well as it is simply a linear combination of the two, right?

The research conducted by Marini et al. presents a novel interpretation of four decades of data from the Solfatara geochemical database, utilizing new geoindicators and aiming to offer potential predictions of future scenarios. The subject matter is of significant scientific relevance, particularly as the Solfatara magmatic-hydraulic system holds considerable interest, not only for the scientific community engaged in the study of this system, but also due to the concerns regarding the potential evolution of the current bradyseism towards eruptive scenarios, which may have substantial societal implications for the population residing in its vicinity. However, despite the relevance of the topic, I feel that the paper in its current state is not ready for publication, and that some concerns about the manuscript should be addressed before publication.

In the following, I would like to point out two reasons I consider to be of great importance, and which should be reconsidered by the authors before a possible publication. Furthermore, in an attached file, I have integrated the comments and suggestions next to the relative text which are aimed at improving the readability of the text and the completeness of the information according to my personal vision.

In summary, the two most important aspects of the manuscript I believe should be reviewed in order to improve the quality of the paper before its possible publication are as follows:

1. The heart of the work consists in the reinterpretation of geochemical data based on the use of new geoindicators. Unfortunately, the text does not contain the information necessary to understand and interpret these geoindicators, which, as the paper is set up, must be well understood to ascertain the fields of applicability and potential limitations. This makes the paper not immediately understandable nor directly usable by the scientific community (after reading this document, can the use of these new geo-indicators be replicated in other cases?), unless one looks for the source, however this is contained in a book that precludes easy acquisition for many. I myself have not been able to find it, as I explained in the attached file. However, the authors of this manuscript are also the authors of the monograph; therefore, it is my opinion that they could easily integrate the text with a specific section describing the new geoindicators, which would certainly enhance the text. As a side effect, I was only partially able to follow the exchange between the first reviewer and the authors, due to the asymmetry of information caused by my lack of knowledge of the content of the monograph cited.

1. Another significant concern relates to the capability to predict specific scenarios following the analysis and model interpretation. I acknowledge that the term "prediction" can have various meanings; however, in scientific literature, it is generally accepted to convey the notion of "the expectation of an occurrence under certain conditions". Typically, this expectation is quantified and qualified through numerical analysis (statistical or probabilistic), which lends credibility to the process of predicting future events or outcomes. Prediction can be viewed as part of hypothesis testing, where one formulates predictions as components of hypotheses. These predictions can then be empirically tested through experiments or observations to confirm or refute the hypotheses. However, this concept is not clearly articulated in the text. Instead, the authors present reasonable scenarios that could evolve in different directions, potentially even oppositely, without providing any arguments or analyses to support specific predictions. The text offers only a description of various possible scenarios, lacking a basis for making predictions. In my opinion, this aspect of the paper needs to be reconsidered, given the significant importance of the ability to make predictions, especially for local authorities responsible for managing risk in a densely populated area like Campi Flegrei. Therefore, I suggest that the authors revise the use of the term "prediction" from the title onward, modifying the text accordingly or providing sufficient justification to support the possibility of predicting potential risk scenarios.

I hope that the critical reading of the manuscript that I propose in its current state can be a constructive stimulus for the authors, regardless of the outcome of the publication.