Mixed Recharge and Epikarst Role in a Complex Metamorphic Karst Aquifer: The Pollaccia System, Apuan Alps (Tuscany, Italy)

Reviewer comment

Reply

Line 46

“…, that are commonly responsible of most of the water storage,..” – while interpretation of water storage in karst to be related to fractures and rock matrix in phreatic zone  is quite common (e.g. your reference 10), there is also growing body of research which results relate storage of karst systems mainly to epikarst and soil zone.

The epikarst can indeed represent an important water storage compartment, depending on the characteristics of the considered karst system (i.e., its relevance is site-specific). There are also examples where its hydrologic role is less relevant. A brief description and its hydrologic significance was added in the Introduction

Lines 49-51

Couldn’t low or high variability of flow and physio-chemical parameters be alternatively explained by efficiency of epikarst/soil zone (mixing and recharge buffering)? Well-developed epikarst with significant soil cover can buffer large variations of water quality at springs, even in systems that have well developed conduit systems in the deeper zones.

It can be the case in karst systems that have well-developed epikarst and soil cover. However, it is not the case for the karst system studied in this work. Only the Panie sub-system has a well-developed epikarst but soil cover is scarce to absent over the whole Pollaccia catchment.

Line 62

You mean primary porosity? As fracturing and conduit systems are clearly present.

We mean both primary and secondary porosities .

Line 171

Can you describe hydrological phases term in more detail – why autumn recharge and spring discharge (both autumn and spring seasons show multiple recharge events)?

In order to simpliphy the concept of hydrological phases we have changed “spring discharge” to “spring recharge”. A description about the hydrological phases definition was added in the Section 3.1.

Line 222

You probably mean Table 3 (not 1)?

Yes, we meant Table 3 (new Table 2), corrected.

Line 391

I do not understand why this suggests drain system with free-surface flow (and not phreatic conduit)? Can you elaborate this conclusion more extensively?

This concept is discussed in detail in the Section 5 of the revised manuscript. The Introduction was also extented with notions that are useful for the discussion.

Lines 451-461

Couldn’t alternate interpretation be that during first part of a winter period you get progressively less contribution from higher parts of the catchment (as EC is generally inversely correlated to the catchment elevation)?

It is an interesting interpretation, however if the suggested phenomenon had occurred, we should also have observed an increase of water temperature. Moreover, the suggested process would have caused increasing EC and T trends until the 19^th of February since relevant infiltration events did not occur until that date.

Lines 397-420

In my opinion additional interpretation of the HTC-graphs shown on Figure 8 should be included in discussion.

The HTC-graphs showed in Fig. 8 are discussed in more detail as suggested.

Line 471

“no epikarst strorage” is exaggerated statement (maybe better to say “less efficient”) – no storage in SSmain is also contrary to your interpretation that during second part of the winter recession contribution from SSmain gets proportionally larger comparing to SSpanie, which is interpreted to have significant epikarst storage.

We agree with the reviewer about the exaggerated statement concerning the epikarst, therefore it was corrected with “reduced epikarst storage”.

Regarding the ss_main,  we did not mean that it has no storage (see former lines 451-461), it is mainly a matter of different flow velocities in the two sub-systems. The Section 5 and the Conclusions were modified to explain better the previous idea.

Generally, karst spring HTC graphs are summed response of all processes that are happening in the karst system (from the soil and epikarst through vadose and phreatic zone to the spring). So extra caution should be required not to oversimplify as several different processes can influence ending results (in similar or opposite direction). E.g. some research is showing that spring water mineralization is directly related to variations of pCO₂ in soil and epikarst during recharge events, without much influence of characteristics and processes in the deeper parts of the system (e.g. Yang et al. 2012).

As general statement, we agree with the reviewer comment. However, two points are relevant for this system:

·       Soil cover is scarce and discontinous in most of the catchment and so does the epikarst.

·       Normally, pCO₂ increases during spring season, therefore we should observe an EC increases. The opposite occurs at the Pollaccia during the main S.D. events.

Reviewer comment

Reply

The title is too long and provides too many details, I suggest the authors consider reducing the length of the title.

The title was shortened as suggested.

Line 142

why the spring monitoring temporal range is 384 days from s (from 2011/11/06 to 2012/11/23)? Due to data availability?

Yes, it is indeed due to data availabilty, an it is also the period in which the data are most continuous.

Line 162

in equation 1, how the constant 0.41 is defined. Please explain more in addition to saying it is an efflux coefficient. The same question applies to other equations. Please explain more on how these equations are developed.

An explanation about the efflux coefficient was added, a detailed discussion on its formulation can be found in any hydraulics textbook. The value of the temperature coefficient in eq. (2) was added. Regarding the storm hydrograph analysis, a new table lists all the parameters and their meaning (new Table 1).

The title says the paper is exploring the roles of recharge and epikarst on hydrodynamic; however, the main body of the paper does not seem align with this topic. In the conclusion there is nowhere mention the impacts of recharge and epikarst. In my opinion. The author should either reorganize the paper a little bit and provide more information in the conclusion section about the experiment results. And the paper should add validation, and comparison with other methods/studies.

The whole paper deals with the roles of recharge and variable epikarst storage on the spring hydrodynamic. The spring monitoring is the starting point of the investigation and the detailed analysis of the HTC-graphs at both the seasonal and single-event scales is the key to unravel the two issues presented in the title. The discussions present our interpretation on how autogenic and allogenic recharge and the epikarst influence the Pollaccia spring behavior. The introduction was expanded to address the topics.

The conclusions were expanded as suggested.

The language should be improved. Grammatical errors and wrong words need to be corrected

The grammatical errors were corrected

In figure 2, please highlight what is the red star and black dot in (a).

Corrected as suggested

Reviewer comment

Reply

Line 20-21

Either explain in the text the meaning of allogenic/autogenic recharge or don’t use them

The meanings of allogenic and autogenic recharge have been added in the Introduction. Anyway, these two terms are fundamental in karst hydrology because they give information on the type of water that infiltrates in the karst system: allogenic water is significantly less mineralized  than autogenic water because fast runoff on poorly soluble terrains do not permit a substantial mineralization.

Line 30

what about permeability?

Added

Line 37

are you presuming your reader will know what you mean by "epikarst"? You should note it here and then explain in detail in your geologic section

The definition and its hydrologic role were added. A brief description of its structure was added in the Section 2.1

Line 43

Somewhere here you need to prepare the reader for Fig. 8 and your discussion of concepts like "piston-flow" or dilution and general principles that control spring temperature and SC. If the reader is thinking about these things your descriptions and data will make more sense.

The definition of all the general principles and concepts were added.

Line 51

diffuse?

The term is a synonym of diffuse and it refers to the definition given by ref. [10].

Line 61

Monitoring of Cl or another anthropogenic contaminant (CFC?) would permit more precise models

We thank the reviewer for the suggestion, this could be useful for future research and is pointed out in the Conclusions section of the revised manuscript.

Line 67 (and elsewhere)

reservoirs? Storage elements? Domains? "Compartment" seems inappropriate.

Changed as suggested in the whole revised manuscript.

Line 75

probably you mean metamorphic

Corrected

Line 89

perhaps you could define even earlier in the text.

Done (see previous replies)

Line 94

these terms are fine for your karst and caves audience, but not for a broader audience you may want to reach.  You really need to define epikarst!  It means different things to different parts of the profession, sort of like the word "saprolite" in some areas of geomorphology

Done (see previous replies)

Line 116

This geologic discussion is interesting for part of your audience, but do you need this much detail  for this paper?  Focus less on the geologic history and more on the materials, including your version of epikarst

The geologic history was shortened and the Materials section was strenghtened as suggested.

Line 132

same comment!

We assume the Reviewer suggests to explain the meaning of vauclusian spring. Its definition is reported in the revised manuscript

Line 135

these would be  mean values?  Maxima?

These are mean values (specified in the revised manuscript). Most of the measurements were done with charcoal detectors, i.e., the detectors are placed at the spring and they are collected and substituted with new ones at specific sampling intervals, according to the meteorological conditions (weekly or after two weeks). The information they give is qualitative rather than quantitative. The results of the only test done with a fluorimeter are reported as a mean flow values for comparison with the other tests. The methods  adopted for the tracer tests (i.e., charcoal and fluorimeter) were added in the text.

Line 156

seeps?  Did you check to see if the estimated Q was correct.  what is h?

It is a small passage that bypasses the weir flowing directly in the main channel and that cannot be measured

Line 169

What is the sampling interval for the precipitation data? The interval is important for all your calculations.

It is 15 minutes but we aggregated the data on a 30 minutes samplig interval to compare the precipitation data with the hydrological data of the Pollaccia spring. Added in the text.

Line 176

Give your reader some sense of scale for the time axis in 3a and 3b.  In some systems the complex patterns you show  in b could result from different Q sources with different lag times, storm motion, etc.   the bulleted points here might be better in a table and you should try and define each of the variables you discuss below. What is delta tp, for instance?

The bulleted point were removed and a new table was added to describe each parameter of the storm HTC-graph (new Table 1). Delta tp  was defined in the old line 190, it is now reported in the new Table 1.

Fig. 3 refers to either a single storm event (i.e., from a few hours to a couple of day) or a multiple one (the storm peaks happen every 0.5 - 1 day) but there is only a meteorological station in this cathment therefore the multiple precipitation peaks show consecutive storm events. The storm peaks are close enough in time to perturb the recession of the previous Q peak but there is not a standard time interval to select as scale. Usually, it is generically named Δt. Individual, consecutive Q peaks are not related to the same storm event as showed by the timing between P and Q peaks.

Line 194

explain what you are "correlating" in a little more detail.  You have both "season", which is probably not a constant , and antecedent ppt, both of which are probably reflected by some measure of GW height.

More information were added in the revised manuscript.

Line 213

This is a really illustrative graph!  If you had to recast it for readers who were red_green colorblind, what would you do?

The suggestion on the colors is well spotted, however each curve is labelled, they are not distinguished by colors only. Moreover, the hydrologic phases are distiguished by both colors and graphic elements (boxes and labels are more highlighted in the revised figure).

Line 221

Explain how you handled the 40 days of unmeasured flow in your average

The Q data was extrapolated in those average because no precipitation occurred therefore an exaustion curve could be modelled. Those data were included in the average Q calculation. EC and T average values were calculated with tha available data.

Line 235

Is there no nearby record of groundwater level?  Record from an observation well would help characterize the state of the system during these "seasonal" periods, as well as the departure from average seasonal values

There are no records of water wells. However, wells give extremely local information on groundwater level in this type of karst aquifers,  it is difficult to use those information at the catchment or sub-catchment scale.

Line 242

Give the source of data

It was given in the methods section but it is reported also in the revised figure caption.

Line 248

How do these values compare to the mean annual air temperature?  Do these springs have a thermal content?

There are not meteorological stations equipped with thermometers for air temperature in the Pollaccia catchment. However, this spring could be considered a “cold” spring because water temperature is surely some °C lower than local air temperature.

Line 250

What produces the sharp temperature spikes apparent in this figure?

These T peaks could reflect input of runoff water in the proximity of the spring. It is discussed in the revised manuscript

Line 256

These are low values for carbonate terrain.  Do you have chemistry or spot chemistry to go with the specific conductance?

We do not have spotted chemical data but these values are common in karst systems that are characterized by high-conductance flow or prevalent free-surface flow.

Line 266 (Table 1 )

Since these values are already in the text or could be, no need for a table!

Table removed

Line 272

In the introduction, perhaps, or perhaps here you need to explain what likely controls SC. It should be a function of bedrock type and contact time since temperature is relatively constant?

Added in the introduction

Line 282 (Table 2)

If you needed to you could show the data collected during your interval and the mean ± 1 sigma since I don't think you need the rest of the data in this table

The Table was removed

Line 299

Pmax is a measure of intensity?  Over what time interval?

It is the maximum precipitation intensity over a 60 min interval (listed in the new Table 1).

Line 301

A useful measure you don't seem to have applied is the storm volume compared to the Qvolume for the same event

We could not measure the volume of the storm because there is only a meteorological station in the catchment. However, this study is focused on the hydrodynamic response rather than on volume estimation. Therefore, time lags and P_max/P_tot/etc./Q_c relationships are more useful for our purposes.

Line 310

Were they both or only positive?

It depends on each storm event. However, T and EC show usually in-phase oscillations (i.e., both positive or negative).

Line 319 (Table 3)

what is this measure??   Need to define in table?

It was described in the old line 190. However, a table with the parameters was added (new Table 1)

Line 322

Did you do a similar analysis for T and SC, or just for the lag time?

We did also for T and EC but no significant correlation were found. However, Q-EC/T lags data were less than the P-Q lags data, therefore we preferred to exclude that analysis to avoid misinterpretations.

Line 328 (Fig. 7)

AR and WDR are not significantly  different, right?

According to the lags, W.Dr. lag values are more dispersed and longer than the A.R. ones

Line 344

What was the ratio of inflow to outflow volumes  for this event?

We could not estimate P volumes (see reply to the comment for Line 190). However, Q_C, Q_C-Q_A, P_tot and Δt_b show that this is a very small event.

Line 351

and “postponed” with “occurred before”

We meant “occurred after”, not “occurred before”.  “Postponed” was replced with “occurred after”

Line 355

Does this reflect the precipitation intensity or initially high values for soil and groundwater storage?

Both, another strong storm occurred few days before and this storm event was particularly intense.

Line 358

This is discussion, but what does it refer to?  The strong recharge event?  All of the examples?   What is small-scale piston flow?  Give a citation since the concept of using new water to push old water out is applied in a variety of ways. You need to set up the discussion in your introduction rather than "spring" it on the reader here--she should know what to expect so that your examples make sense

Added (see the previous comments)

Line 360

What would its chemistry and temperature be?

Runoff water would be less mineralized than groundwater (GW). Its temperature can be higher or lower than GW, depending on climate, meteorological, and GW flow conditions.

Line 364

More generally--air and near-surface temperature vs GW temperature.  This relationship will have an annual cycle and/or there may be an observation well that records it nearby.

There are no wells (see previous comments). However, the relationship between seasonal air T trends and the spring hydrodynamic response at the storm-scale is not straightforward.

Line 380 (Table 5)

Should also have columns for the temperature and SC response so your readers can see what is in phase

Added as suggested.

Line 384

How does climate influence--you demonstrate that season is important....and weather

Some notes were added.

Line 387

this refers to hydrogeologic environments? Your response times are almost what you'd expect for open-channel flow?

This refers to the recharge type, which is dependent on the the geological-geomorphological setting.

Line 389

Table 5 should show this. 2/3 of your examples illustrate the opposite

The first example illustrates a piston, the second one shows a strong dilution, the third one is a mixed event with a clear dilution during the flood peak. EC trend in Fig. 4 and new Table 4 show that most of the events are dilutions. Added two columns to Table 4 for T and EC .

Line 405

Change ”precipitations were” and “than those occurred” with “precipitation was” and “than events that occurred”, respectively. Does “impulsive” mean more intense?

Impulsive means more intense and more concetrated in a short time interval, rewritten as “intense”.

Line 412

“postponing” do you mean “occurring before”?

We mean the opposite, i.e. occurring after (see a previuos comment)

Line 418

are they catchments?

They are parts/domains of the Pollaccia spring catchment, therefore they are named either “sub-systems” or “sub-catchments”. The former definition was adopted in the whole revised manuscript

Line 418-419

But with piston flow coming from each domain?  You need to better explain the key observation of temperature and SC (sometimes) spikes on rising limbs

Explanations added

Line 421-461 (last part of the Discussions)

This section deserves its own subheader here and a setup in the introduction for the visual model you show here that includes components with different residence/travel times and different chemistry.  Then you need to organize and tighten your explanation, referring the reader back to new information you are going to add to the introduction about the chemistry of and residence time of water in the epikarst.  You shouldn't be introducing this detailed description of materials,structures in the discussion.  You should be explaining how the material properties lead to what you observed and portray in your model.  Your explanation seems mainly reasonable, but your writing needs to be much simpler and more clear for this important section

Rearranged as requested.

Line 425

Drainage?

We refer to different discharges as explained in the following rows.

Lines 442-448

All of this appropriate iinformation belongs with an earlier description of epikarst

The description and hydrologic role of the epikarst were added in the Introduction section of the revised manuscript. However, these information are useful in this section because they give additional proofs on the model that we propose.

Line 461

This seems plausible and consistent with this part of the dataset. Is it consistent with the rest of the data. Would the epikarst produce a warm water spike on the rising limb at the end of February?

The epikarst water cannot produce a warm water spike on the rising limb at the end of February because its response is a long-time lasting contribution. As discussed in Section 5, the T/EC spike that was observed in that period is related to water stored in the phreatic zone.

Figure 9

Spell it the same way each time, if you really insist on using it! [comment on “autigenic” recharge]

Corrected with “autogenic” as in the whole manuscript

Line 470

Try deleting allogenic and authigenic here and elsewhere.  Do the words really add to what you are saying?

As discussed before, the terms autogenic and allogenic (recharge) have specific meanings in karst hydrology that are useful for the purposes of our study

Line 480

[referring to “compartments”] It'd be good if you had a different word to describe these contributing areas--domain would be better and I am sure you can think of something more descriptive

Changed as suggested in the whole revised manuscript.