Hydrogeochemical Study on Closed-Basin Lakes in Cold and Semi-Arid Climates of the Valley of the Gobi Lakes, Mongolia: Implications for Hydrology and Water Chemistry of Paleolakes on Mars

Round 1

Reviewer 1 Report

Dear Editor and Authors,

 

Please find below my review of the article “Hydrogeochemical study on closed-basin lakes in cold and semi-arid climates of the Valley of the Gobi Lakes, Mongolia: Implications for hydrology and water chemistry of paleolakes on Mars”, submitted to Minerals by Y. Sekine and coauthors.

The article describes the hydrogeology, aqueous geochemistry and mineralogy of a semi-arid closed basin lake system, the Valley of the Gobi Lakes in Mongolia. Based on the analyses of data collected over two seasons of fieldwork, the paper proposes that in addition to precipitation and meltwater runoff, groundwater originating from two different sources plays an important role in dictating the geochemistry of the lakes present. The paper draws comparisons to Early Mars evaporite-bearing paleolakes, and suggests that diverse forms of groundwater flow may likewise have played a role in the formation and geochemistry of the Early Mars paleolakes, given an early arid/semi-arid climate.

Overall, I found the manuscript to be very interesting, well written and scientifically relevant. To my knowledge, the Valley of the Gobi Lakes has never been explored as a planetary analog to Early Mars lakes and hydrology. My recommendation is for the manuscript to be published, subject to minor corrections and suggested edits to improve the clarity of the manuscript and strengthen the arguments presented.

Particularly, the paper did not sufficiently describe similarities and differences with other previously suggested terrestrial analogs for Early Mars lake systems. A small section in the Discussion showing how the site’s hydrology, geochemistry and other characteristics differ from other cold semi-arid to hyperarid basin lakes on Earth proposed as Early Mars lake analogs (e.g., Qaidam Basin, Atacama and Andean lakes, Axel Heiberg Island cold springs) would be very informative. I suggest a non-exhaustive list of relevant papers below that may be helpful to carry out this task.

While not strictly necessary for publication, adding further information about the minerals that control the composition of the lake, river and groundwater would make the investigation more compelling. Particularly, providing this information could address the question about the residence time of groundwater fluids and would therefore improve the interpretation of the hydrology. The conclusions suggest that a mixture of groundwater fluids with distinct compositions would be required to explain the lake chemistry, and that one of these groundwaters (“Group A”, containing relatively high bicarbonate concentration) may originate from river water (not as much carbonate) equilibrating with soil and subsurface carbonate minerals. However, the origin of the carbonate minerals is not fully described. Are they detrital? Formed by weathering and precipitation? Pedogenic? Additionally, the timescale of equilibration with subsurface carbonate can be calculated with known reaction rate constants (e.g., Palandri and Kharaka 2004; A compilation of rate parameters of water-mineral interaction kinetics for application to geochemical modeling) to constrain a minimum fluid residence/transport time, and this may help identify whether groundwater sources are indeed important contributors to the lakes. I offer further details below.

 

Detailed review:

Throughout the manuscript: it would be preferable to include the charge of the HCO₃⁻ , SO₄²⁻ and CO₃²⁻ ions in the text, out of convention (instead of HCO₃, SO₄ and CO₃).

Throughout the manuscript: The º symbol was used instead of the degree (°) symbol for temperatures.

Line 75: evaporation predominates precipitation -> evaporation predominates over precipitation.

Line 77: Are the Khangai Mountains really the southernmost edge of permafrost occurrence in the Northern hemisphere? What about the Himalayas? Maybe you mean “continuous permafrost”?

Line 108: finish in the Böön Tsagaan and Orog lakes -> end in the Böön Tsagaan and Orog lakes.

Line 110: “depths of 10 m and 3 m”. Are these depths the mean depths or maximum depths? If these are mean depths, one can calculate whether the catchments really do not receive sufficient precipitation to replenish the lakes yearly (Lines 133 – 134). For Böön Tsagaan (250 km² (line 109) * 10 m (line 110) = 2.5*10⁹ m³ water), assuming a precipitation of 50 – 100 mm (Line 126) * 45000 km² (line 111) = 2.25*10⁹ m³ – 4.5*10⁹ m³, so in fact, average yearly precipitation can sometimes supply sufficient water to maintain the same lake volume at Böön Tsagaan. For Orog (140 km² (line 109) * 3 m (line 110) = 4.2*10⁸ m³ w ater), 50 – 100 mm (Line 126) * 14900 km² (line 111) = 7.45*10⁸ m³ – 1.49*10⁹ m³, so on average there would be enough yearly precipitation in the Tuin catchment to fill Orog Lake; no extra groundwater would be required. Evaporation rates would change these numbers and whether the average precipitation would be sufficient to fill the lakes, so an estimation of the open water evaporation rate by the energy budget would be necessary. Are there measurements for the amounts of discharge of the rivers into the lakes?

Line 111: Inconsistent use of comma in “~45000 km²” and “14,900 km²”.

Figure 2: It would be very interesting and informative to see where the catchment divide line is on this map, between the Baidrag and Tuin catchments. Which catchment does the Tsagaan Nuur wetland belong to? That could inform whether groundwater originating from the wetland ends up predominantly in one lake or the other.

Line 186: portable electrical portable pH meter -> portable electrical pH meter.

Line 199: the REACT in Geochemist’s Workbench -> the REACT program in Geochemist’s Workbench.

Line 201: the REACT in Geochemist’s Workbench -> the REACT program in Geochemist’s Workbench.

Line 203: through removing water abundance -> by progressively removing pure water.

Lines 201 – 204: At what temperature was the evaporation modeled?

Lines 237 – 238: Please provide the chemical formulae for MHC and AMC here, where they are first mentioned.

Table 1: (2018 Samples) Are data for BTS well 2 missing in the table? Also, line 192 says Si concentrations were also measured, but they are not shown in this table.

Figures 3 and 4: There is a typo in the caption: Are the units meq/kg or meg/kg?

Line 300: bts -> BTS.

Section 4.2. Was it not possible to quantify the relative volumes of the minerals identified by XRD? Quantitative XRD results could help with the evaporation models carried out with GWB. For example, it may be interesting to compare the abundance of minerals precipitated in the GWB evaporation models to the minerals identified by XRD to see validate the evaporation models. This could provide an additional or parallel criterion to understand whether evaporation leads to the measured lake water composition.

Line 372: Since Figure 5 contains the word “smectite” in the legend instead of saponite, either change the legend to say “saponite”, or change the first appearance of “saponite” (Line 372) to “saponite (a smectite)”.

Discussion: I would have liked to see a few sentences or a short section comparing and contrasting the results of this analog site to other cold, semi-arid to hyperarid analog sites suggested to represent lake systems on early Mars. Particularly, the Qaidam Basin, Atacama and Andean lakes, and Axel Heiberg Island cold springs come to mind. Some papers about those sites are:

• Xiao, L.; Wang, J.; Dang, Y.; Cheng, Z.; Huang, T.; Zhao, J.; Xu, Y.; Huang, J.; Xiao, Z.; Komatsu, G. A new terrestrial analogue site for Mars research: The Qaidam Basin, Tibetan Plateau (NW China). Earth-Science Reviews 2017, 164, 84–101, doi:10.1016/j.earscirev.2016.11.003.
• Anglés, A.; Li, Y. The western Qaidam Basin as a potential Martian environmental analogue: An overview. Journal of Geophysical Research: Planets 2017, 122, 856–888, doi:10.1002/2017JE005293.
• Battler, M.M.; Osinski, G.R.; Banerjee, N.R. Mineralogy of saline perennial cold springs on Axel Heiberg Island, Nunavut, Canada and implications for spring deposits on Mars. Icarus 2013, 224, 364–381, doi:10.1016/j.icarus.2012.08.031.
• Astrobiology Special Collection: A New Mars Analog Site on the Third Pole of Earth: Dalangtan Saline Playa in the Hyperarid Qaidam Basin on Tibet Plateau (https://www.liebertpub.com/toc/ast/18/10).
• Risacher, F.; Alonso, H.; Salazar, C. The origin of brines and salts in Chilean salars: a hydrochemical review. Earth-Science Reviews 2003, 63, 249–293, doi:1016/S0012-8252(03)00037-0.

Line 415 – 418: Please clarify here or in the Methods section whether any minerals were inhibited in the GWB evaporation model (e.g., because of kinetics). For example, magnesite precipitation is kinetically inhibited at low temperature – was magnesite one of the minerals that could precipitate in the model?

Line 422: found the presence of MHC in suspended materials -> found MHC present in the suspended materials.

Line 423 – 427: The sulfate ion was assumed to be in solution and not precipitating in the GWB evaporation model, and the rationale is that the fluid composition (Table 1) is far from the saturation of gypsum. However, were other low temperature sulfate minerals considered? For example, the solution may be closer to saturation of mirabilite or epsomite (e.g., Chou et al. 2013; The stability of sulfate and hydrated sulfate minerals near ambient conditions and their significance in environmental and planetary sciences. Journal of Asian Earth Sciences, 62, 734–758, doi:10.1016/j.jseaes.2012.11.027; Wang et al. 2016, Setting constraints on the nature and origin of the two major hydrous sulfates on Mars: Monohydrated and polyhydrated sulfates. Journal of Geophysical Research: Planets, 121, 678–694, doi:10.1002/2015JE004889).

Line 433 – 434: mx represents an activity of components X in a solution -> mx represents the activity of component X in solution.

Line 438: precipitations -> precipitation.

Line 441: lake water, suggesting that the contributions of groundwater are -> lake water composition, suggestion that the contribution of groundwater is.

Line 461 – 462: Is the HCO3 and CO3 high because the water is in equilibrium with a 4*10^-4 atm pCO2 atmosphere, not because of the dissolution of MHC and AMC?

Line 500: would -> could.

Line 501: ones -> lakes.

Line 502: would -> could.

Line 499 – 504: The water mass balance discussion would be more compelling and convincing if values for the discharge into the lakes were included. Are the amounts of discharge known? What about evaporation?

Line 510: explain the lake water -> explain the lake water composition.

Line 515 – 518: Please clarify: by “occurrence of permafrost in the mountains” influencing the aqueous organic compositions, do you specifically mean that permafrost thawing/melting controls the aqueous organic abundances?

Line 518 – 520: This sentence is a little confusing. Perhaps it would be clearer if it was rephrased. Here is a suggestion: “To explain the conflicting results that the lake composition derives from group-A groundwater and yet the lake’s organic composition is influenced by permafrost thaw in the Khangai mountains, we suggest that the source of the group A groundwater could be river water that infiltrated the subsurface.”

Line 520: considerable -> significant.

Line 522: During river water’s infiltration: During the river water’s infiltration.

Line 522 – 525: But what is the ultimate source of the calcite concretions? Adding information about the origin of the calcite would be of interest and could strengthen the hypothesis that the calcite controls the carbonate and Ca concentration of the groundwater and ultimately the lake water. Specifically, consider calculating a plausible residence time to find out if the transport from the groundwater sites to the lake occur on reasonable timescales. The Baidrag river would have to pick up 1 meq alkalinity/liter (BTS South well, Table 1), during its groundwater stage by equilibration with groundwater calcite. The dissolution rate of calcite can be found in Palandri and Kharaka (2004; compilation of rate parameters of water-mineral interaction kinetics for application to geochemical modeling; Geological Survey Menlo Park CA). Alternatively, the rate can be calculated with GWB.

Line 530 – 532: This would be a stronger argument if saturation for other sulfate salts were also considered.

Line 536 – 537: saline water infiltrated the subsurface -> saline water may have infiltrated the subsurface.

Line 549 – 553: Is the source of the water of the Tsagaan Nuur wetland known? Is the volume of the water in the wetland known?

Figure 8: This schematic is a very useful summary! But it brings to mind: could overland flow supply water to the Böön Tsagaan lake, in addition to (or instead of) groundwater? There appears to be a topographic barrier between the Tsagaan Nuur wetland and the Böön Tsagaan lake in this figure, but It is hard to appreciate if there is a topographic barrier n Figure 2. It would be helpful to plot the topographic barrier on Figure 2.

Line 600: diageneted -> diagenetic.

Line 641 – 648: What would the source of water at the Tsagaan Nuur wetland be? Would it also be fed by the Baidrag river, or overland flow? Why would evaporation of water at the Tsagaan Nuur wetland lead to water so much more saline than at the lakes?

Author Response

Responses to comments from Reviewer #1 for ‘Hydrogeochemical study on closed-basin lakes in cold and semi-arid climates of the Valley of the Gobi Lakes, Mongolia: Implications for hydrology and water chemistry of paleolakes on Mars’ by Sekine, Y., et al.

 

We are grateful to reviewer #1 for the constructive comments and suggestions. Below, we give our responses in turn following each comment, with the reviewers’ comments being in Arial font and our responses being in Times font. 

Major Comment 1: Particularly, the paper did not sufficiently describe similarities and differences with other previously suggested terrestrial analogs for Early Mars lake systems. A small section in the Discussion showing how the site’s hydrology, geochemistry and other characteristics differ from other cold semi-arid to hyperarid basin lakes on Earth proposed as Early Mars lake analogs (e.g., Qaidam Basin, Atacama and Andean lakes, Axel Heiberg Island cold springs) would be very informative. I suggest a non-exhaustive list of relevant papers below that may be helpful to carry out this task.

 

Response to Major Comment 1

We greatly appreciate the reviewer’s suggestion and providing the list of the relevant papers. A major difference of the Valley of the Gobi Lakes from other terrestrial analogues is the sources of water and salinity to lake systems. In Atacama-Andes regions, there are multiple sources of salts to the lake systems, including atmospheric transport of sea salts, desert dusts, recycling of crustal salts, and volcanic inputs. This is because the Andes is a tectonically-active and hyperarid area. The major source of saline water to Axel Heiberg Island springs is subglacial meltwater. This happens because Axel Heiberg Island locates in polar zones where thick glaciers can grow. In a cold and semi-arid area of the Valley of the Gobi Lakes, seasonal melting of ground ice and its infiltration into soils provide water and salinity. We consider that the present study could provide a simplified view of groundwater hydrology. The common characteristics among these analogue sites are the fact that groundwater activities play central roles in hydrology.

To explain other terrestrial analogues for early Mars lakes, we have added the following sentences in Introduction, in lines 67–71 of the revised manuscript:

“Nevertheless, only a few studies have investigated saline lacustrine environments that developed in cold and semi-arid climates as terrestrial analogues for early Martian lakes, such as cold springs on Axel Heiberg Island in the polar zone [22], saline lakes on the Qaidam Basin in the arid climate zone [23], and salt lakes in the Atacama-Andes regions [24]. In especially, only few studies have investigated subsurface hydrogeochemical cycles and roles of groundwater.”

According to the comment, we have also added the following sentences in Discussion, in lines 630–640 of the revised manuscript.

“Comparing to other terrestrial analogues for early Martian lakes (e.g., saline lakes in the Qaidam Basin and Atacama-Andes regions, and cold springs on Axel Heiberg Island) [22–24], the sources of water and salts differ each other depending on the geohydrological settings. Owing to the tectonically-active setting and hyperarid climate of Atacama-Andes regions, there are multiple sources of salts, including aeolian inputs (sea salts and desert dust), recycling of crustal salts, and volcanic inputs [24]. In the polar zone of Axel Heiberg Island springs [22], the major source of highly saline water is subglacial meltwater. Although these supply processes of salts could have also occurred in lake systems on early Mars, the characteristics of the Valley of the Gobi Lakes allow us to understand the roles of evaporative concentration of salinity and groundwater hydrology in semi-arid climates. Despite of the variety of the water sources among the analogue sites, all of them suggest the importance of groundwater hydrology to transport the water and dissolved elements to the lakes.”

 

 

Major Comment 2: While not strictly necessary for publication, adding further information about the minerals that control the composition of the lake, river and groundwater would make the investigation more compelling. Particularly, providing this information could address the question about the residence time of groundwater fluids and would therefore improve the interpretation of the hydrology. The conclusions suggest that a mixture of groundwater fluids with distinct compositions would be required to explain the lake chemistry, and that one of these groundwaters (“Group A”, containing relatively high bicarbonate concentration) may originate from river water (not as much carbonate) equilibrating with soil and subsurface carbonate minerals. However, the origin of the carbonate minerals is not fully described. Are they detrital? Formed by weathering and precipitation? Pedogenic? Additionally, the timescale of equilibration with subsurface carbonate can be calculated with known reaction rate constants (e.g., Palandri and Kharaka 2004; A compilation of rate parameters of water-mineral interaction kinetics for application to geochemical modeling) to constrain a minimum fluid residence/transport time, and this may help identify whether groundwater sources are indeed important contributors to the lakes.

 

Response to Major Comment 2

Given the occurrence of spherical carbonate concretions in the soils (Supplementary Fig. 1), they were likely formed through weathering and precipitation of calcite-saturated porewater within the sediments during early diagenesis (Yoshida et al., 2018, Science Adv. [18]). This has been added in lines 581–586 of the revised manuscript as follows:

“The calcite concretions may have been generated through early diagenesis of carbonate-containing sediments of a larger lake in the past [18]. As described above in Sec. 1, the Valley of the Gobi Lakes would have been once covered with a larger lake in more humid past [28]. We found bedding of the lacustrine sediments in the outcrops (Supplementary Fig. 1). We suggest that early diagenesis of the carbonate-containing sediments might be a source of calcite concretion in the soils.”

We would also appreciate the reviewer’s suggestion to estimate a timescale of equilibrium of carbonate dissolution using the measured dissolution rate; however, we did not measure the surface areas of carbonate minerals in the soils. Accordingly, an estimation of timescale totally depends on the assumption of the surface area. Although we consider that the reviewer’s suggestion is interesting and important, we consider that this could be a future work of the present study.

 

 

Detailed Comments:

Comment 3: Throughout the manuscript: it would be preferable to include the charge of the HCO₃⁻ , SO₄²⁻ and CO₃²⁻ ions in the text, out of convention (instead of HCO₃, SO₄ and CO₃).

 

Response to Comment 3: This has been revised throughout the manuscript.

 

 

Comment 4: Throughout the manuscript: The º symbol was used instead of the degree (°) symbol for temperatures.

 

Response to Comment 4: This has been modified throughout the manuscript.

 

 

Comment 5: Line 75: evaporation predominates precipitation -> evaporation predominates over precipitation.

 

Response to Major Comment 5: This has been modified.

 

 

Comment 6: Line 77: Are the Khangai Mountains really the southernmost edge of permafrost occurrence in the Northern hemisphere? What about the Himalayas? Maybe you mean “continuous permafrost”?

 

Response to Comment 6: To avoid confusion, this description has been deleted.

 

 

Comment 7: Line 108: finish in the Böön Tsagaan and Orog lakes -> end in the Böön Tsagaan and Orog lakes.

 

Response to Comment 7: This has been modified.

 

 

Comment 8: Line 110: “depths of 10 m and 3 m”. Are these depths the mean depths or maximum depths? If these are mean depths, one can calculate whether the catchments really do not receive sufficient precipitation to replenish the lakes yearly (Lines 133 – 134). For Böön Tsagaan (250 km2 (line 109) * 10 m (line 110) = 2.5*109 m3 water), assuming a precipitation of 50 – 100 mm (Line 126) * 45000 km2 (line 111) = 2.25*109 m3 – 4.5*109m3, so in fact, average yearly precipitation can sometimes supply sufficient water to maintain the same lake volume at Böön Tsagaan. For Orog (140 km2 (line 109) * 3 m (line 110) = 4.2*108 m3 w ater), 50 – 100 mm (Line 126) * 14900 km2 (line 111) = 7.45*108 m3 – 1.49*109 m3, so on average there would be enough yearly precipitation in the Tuin catchment to fill Orog Lake; no extra groundwater would be required. Evaporation rates would change these numbers and whether the average precipitation would be sufficient to fill the lakes, so an estimation of the open water evaporation rate by the energy budget would be necessary. Are there measurements for the amounts of discharge of the rivers into the lakes?

 

Response to Comment 8: We appreciate the reviewer’s estimates. As suggested by the reviewer, the open water evaporation rate is important to evaluate whether the precipitation can be enough to explain the lake water volume. The aridity index (= the ratio of precipitation to potential evaporation) of the Valley of the Gobi Lakes is reported as 0.2–0.3 (Davaa et al., 2006). Thus, the open water evaporation is three to five times the precipitation. We have added the following sentence in lines 151–152 of the revised manuscript.

“The aridity index (the ratio of precipitation to potential evaporation) is estimated to be 0.2–0.3 in the Valley of the Gobi Lakes [25].”

 

 

Comment 9: Line 111: Inconsistent use of comma in “~45000 km2” and “14,900 km2”

 

Response to Comment 9: This has been modified.

 

Comment 10: Figure 2: It would be very interesting and informative to see where the catchment divide line is on this map, between the Baidrag and Tuin catchments. Which catchment does the Tsagaan Nuur wetland belong to? That could inform whether groundwater originating from the wetland ends up predominantly in one lake or the other.

 

Response to Comment 10: The Tsagaan Nuur wetland belongs to the catchment of the Bӧӧn Tsagaan Lake according to the previous study (Szuminska, 2016). The addition of the catchment divide line in the satellite image is technically difficult for us. Instead, we described the Tsagaan Nuur wetland belongs to the catchment area of the Bӧӧn Tsagaan Lake in lines 121–122 of the revised manuscript, as follows.

“The Tsagaan Nuur wetland belongs to the catchment area of the Böön Tsagaan lake [26].”

 

 

Comment 11: Line 186: portable electrical portable pH meter -> portable electrical pH meter.

 

Response to Comment 11: This has been modified.

 

 

Comment 12: Lines 199 and 201: the REACT in Geochemist’s Workbench -> the REACT program in Geochemist’s Workbench.

 

Response to Comment 12: This has been modified.

 

 

Comment 13: Line 203: through removing water abundance -> by progressively removing pure water.

 

Response to Comment 13: This has been modified as the reviewer suggested.

 

 

Comment 14: Lines 201 – 204: At what temperature was the evaporation modeled?

 

Response to Comment 14: The temperature was set to be 20 °C. We have added “at temperature of 20 °C” in line 228 of the revised manuscript.

 

 

Comment 15: Lines 237 – 238: Please provide the chemical formulae for MHC and AMC here, where they are first mentioned.

 

Response to Comment 15: This has been done.

 

 

Comment 16: Table 1: (2018 Samples) Are data for BTS well 2 missing in the table? Also, line 192 says Si concentrations were also measured, but they are not shown in this table.

 

Response to Comment 16: We unfortunately did not perform water sampling at BTS well 2 in 2018 due to limitation of time in the field survey. Since we measured Si concentrations only for 2018, we have deleted the description of the measurement of Si from the revised manuscript.

 

 

Comment 17: Figures 3 and 4: There is a typo in the caption: Are the units meq/kg or meg/kg?

 

Response to Comment 17: We thank the reviewer for the careful check. We have corrected the typo. The unit of meq/kg is correct.

 

 

Comment 18: Line 300: bts -> BTS.

 

Response to Comment 18: This has been modified.

 

 

Comment 19: Section 4.2. Was it not possible to quantify the relative volumes of the minerals identified by XRD? Quantitative XRD results could help with the evaporation models carried out with GWB. For example, it may be interesting to compare the abundance of minerals precipitated in the GWB evaporation models to the minerals identified by XRD to see validate the evaporation models. This could provide an additional or parallel criterion to understand whether evaporation leads to the measured lake water composition.

 

Response to Comment 19: We agree that quantitative data on mineral compositions would be helpful to compare with the model calculations; however, we unfortunately cannot quantify the mineralogical compositions using our available software of XRD analysis. The abundances of amorphous phase and/or organic matter are also challenging to the quantification. Although we appreciate the reviewer’s constructive suggestion, we consider that quantification of minerals will be a future work of the present study. Instead we have created a new Supplementary Table 1 to compare the major minerals having XRD intensities > 2000.

 

 

Comment 20: Line 372: Since Figure 5 contains the word “smectite” in the legend instead of saponite, either change the legend to say “saponite”, or change the first appearance of “saponite” (Line 372) to “saponite (a smectite)”.

 

Response to Comment 20: We appreciate the reviewer’s careful check. We have deleted “saponite” from the revised manuscript.

 

 

Comment 21: Discussion: I would have liked to see a few sentences or a short section comparing and contrasting the results of this analog site to other cold, semi-arid to hyperarid analog sites suggested to represent lake systems on early Mars. Particularly, the Qaidam Basin, Atacama and Andean lakes, and Axel Heiberg Island cold springs come to mind.

 

Response to Comment 21: This has been added in the revised manuscript (see our Response to Major Comment 1 above).

 

 

Comment 22: Line 415 – 418: Please clarify here or in the Methods section whether any minerals were inhibited in the GWB evaporation model (e.g., because of kinetics). For example, magnesite precipitation is kinetically inhibited at low temperature – was magnesite one of the minerals that could precipitate in the model?

 

Response to Comment 22: Since AMC and MHC are the initial phases of Mg and Ca-carbonates, magnesite and calcite are practically suppressed in our calculations. We have added the following sentence in line 469– 471 of the revised manuscript.

“Since MHC and AMC are considered to be the initial phases of precipitation of Mg- and Ca-bearing carbonate minerals [35,36], formations of magnesite and calcite are suppressed in our calculations.”

 

 

Comment 23: Line 423 – 427: The sulfate ion was assumed to be in solution and not precipitating in the GWB evaporation model, and the rationale is that the fluid composition (Table 1) is far from the saturation of gypsum. However, were other low temperature sulfate minerals considered? For example, the solution may be closer to saturation of mirabilite or epsomite (e.g., Chou et al. 2013; The stability of sulfate and hydrated sulfate minerals near ambient conditions and their significance in environmental and planetary sciences. Journal of Asian Earth Sciences, 62, 734–758, doi:10.1016/j.jseaes.2012.11.027; Wang et al. 2016, Setting constraints on the nature and origin of the two major hydrous sulfates on Mars: Monohydrated and polyhydrated sulfates. Journal of Geophysical Research: Planets, 121, 678–694, doi:10.1002/2015JE004889).

 

Response to Comment 23: The water chemistry of evaporating lake is also far from the saturation of mirabilite or epsomite. We have added the following expression in line 476 of the revised manuscript.

“This is due to the concentrations of these elements being far below the saturation of chloride and sulfate minerals, including gypsum, epsomite, and mirabilite, in the lake (Tables 2 and 3).”

 

 

Comment 24: Line 433 – 434: mx represents an activity of components X in a solution -> mx represents the activity of component X in solution.

 

Response to Comment 24: This has been modified.

 

 

Comment 25: Line 438: precipitations -> precipitation.

 

Response to Comment 25: This has been modified.

 

 

Comment 26: Line 441: lake water, suggesting that the contributions of groundwater are -> lake water composition, suggesting that the contribution of groundwater is.

 

Response to Comment 26: This has been modified.

 

 

Comment 27: Line 461 – 462: Is the HCO3 and CO3 high because the water is in equilibrium with a 4*10^-4 atm pCO2 atmosphere, not because of the dissolution of MHC and AMC?

 

Response to Comment 27: The high concentrations of HCO₃ and CO₃ are because of both dissolution of MHC and AMC and equilibrium with 400 ppm pCO₂ in alkaline solution. We have modified the sentence in lines 511–512 of the revised manuscript, as follows.

“In contrast, the concentrations of HCO₃^- and CO₃^2- remain relatively high owing to both equilibrium with 400 ppm of atmospheric CO₂ in alkaline solution and following the dissolution equilibria of MHC and AMC.”

 

 

Comment 28: Line 500 and 502: would -> could.

 

Response to Comment 28: This has been modified as suggested.

 

 

Comment 29: Line 501: ones -> lakes.

 

Response to Comment 29: This has been modified.

 

 

Comment 30: Line 499 – 504: The water mass balance discussion would be more compelling and convincing if values for the discharge into the lakes were included. Are the amounts of discharge known? What about evaporation?

 

Response to Comment 30: As far as we investigated, there are no data on the discharge rates for the lakes. This may be because there are no city or town near the lakes. Although we cannot add the data of discharge rate, we have added the information on potential evaporation in the revised manuscript (see Response to Comment 8 above).

 

 

Comment 31: Line 510: explain the lake water -> explain the lake water composition.

 

Response to Comment 31: This has been modified as the reviewer suggested.

 

 

Comment 32: Line 515 – 518: Please clarify: by “occurrence of permafrost in the mountains” influencing the aqueous organic compositions, do you specifically mean that permafrost thawing/melting controls the aqueous organic abundances?

 

Response to Comment 32: Szopińska et al. (2016) reported that organic compounds found in the water column of the Bӧӧn Tsagaan lake are largely same as those found in melting water of permafrost in the Khangai mountains. According to the reviewer’s comment, we have added the following sentence to explain this in lines 570–572 of the revised manuscript.

“The previous study showed that organic compounds bound in water of the Bӧӧn Tsagaan lake are derived from melting water of permafrost in the Khangai mountains [27].”

 

 

Comment 33: Line 518 – 520: This sentence is a little confusing. Perhaps it would be clearer if it was rephrased. Here is a suggestion: “To explain the conflicting results that the lake composition derives from group-A groundwater and yet the lake’s organic composition is influenced by permafrost thaw in the Khangai mountains, we suggest that the source of the group A groundwater could be river water that infiltrated the subsurface.”

 

Response to Comment 33: We appreciate the suggestion by the reviewer. We have rephrased the sentence following the reviewer’s suggestion.

 

 

Comment 34: Line 520: considerable -> significant.

 

Response to Comment 34: To express that the above-mentioned mechanism can happen, we have modified the sentence in line 575 as follows.

“This can happen because the groundwater of group A is distributed downstream of the Baidrag river near the Bӧӧn Tsagaan lake (Fig. 2b).”

 

 

Comment 35: Line 522: During river water’s infiltration: During the river water’s infiltration.

 

Response to Comment 35: This has been modified.

 

 

Comment 36: Line 522 – 525: But what is the ultimate source of the calcite concretions? Adding information about the origin of the calcite would be of interest and could strengthen the hypothesis that the calcite controls the carbonate and Ca concentration of the groundwater and ultimately the lake water. Specifically, consider calculating a plausible residence time to find out if the transport from the groundwater sites to the lake occur on reasonable timescales. The Baidrag river would have to pick up 1 meq alkalinity/liter (BTS South well, Table 1), during its groundwater stage by equilibration with groundwater calcite. The dissolution rate of calcite can be found in Palandri and Kharaka (2004; compilation of rate parameters of water-mineral interaction kinetics for application to geochemical modeling; Geological Survey Menlo Park CA). Alternatively, the rate can be calculated with GWB.

 

Response to Comment 36: The calcite concretions may be generated through early diagenesis of the carbonate-containing sediments (Yoshida et al., 2018 [18]). Yoshida et al. (2018) investigated the formation and alteration of calcite concretion in Mongolia and Utah, USA. They suggested that calcite concretion can be originally generated through early diagenesis of carbonate-containing sediments. In fact, there was a larger lake in the Valley of the Gobi Lakes in more humid climates in the past [25]. The soils near the Gobi lakes were lakebed of the larger lake in the past (bedding of the lacustrine sediments can be seen in Supplementary Fig. 1). To explain the possibility of the origin of calcite concretions, we have added the following sentences in lines 581–586 of the revised manuscript.

“The calcite concretions may have been generated through early diagenesis of carbonate-containing sediments of a larger lake in the past [18]. As described above in Sec. 1, the Valley of the Gobi Lakes would have been once covered with a larger lake in more humid past [28]. We found bedding of the lacustrine sediments in the outcrops (Supplementary Fig. 1). We suggest that early diagenesis of the carbonate-containing sediments might be a source of calcite concretion in the soils.”

 

 

Comment 37: Line 530 – 532: This would be a stronger argument if saturation for other sulfate salts were also considered.

 

Response to Comment 37: We confirm that the water chemistry of Tsagaan Nuur is unsaturated with epsomite or mirabilite. We have added the following sentence in lines 593– 595 of the revised manuscript.

“The near zero saturation indexes of gypsum and calcite suggest that the Ca²⁺, Mg²⁺, HCO₃^-, and SO₄^2- are controlled by dissolution of these secondary minerals in Tsagaan Nuur wetland; whereas, the water chemistry of Tsagaan Nuur wetland is unsaturated with epsomite or mirabilite (e.g., the saturation indexes of epsomite and mirabilite are -1.62 and -1.48, respectively, at 25 °C).”

 

 

Comment 38: Line 536 – 537: saline water infiltrated the subsurface -> saline water may have infiltrated the subsurface.

 

Response to Comment 38: This has been modified.

 

 

Comment 39: Line 549 – 553: Is the source of the water of the Tsagaan Nuur wetland known? Is the volume of the water in the wetland known?

 

Response to Comment 39: The water source of Tsagaan Nuur wetland includes precipitation in a rainy season. The volume of the water is largely fluctuated (e.g., almost zero in a dry season) depending on seasons and years. We have added the following expression in lines 616– 618 of the revised manuscript.

“The groundwater of group B-east would be generated through recharge of highly saline surface water into the subsurface near Tsagaan Nuur wetland (Fig. 8) , whose water source includes small transient rivers from the mountain area and flooding of the Baidrag river in rainy seasons.”

 

 

Comment 40: Figure 8: This schematic is a very useful summary! But it brings to mind: could overland flow supply water to the Böön Tsagaan lake, in addition to (or instead of) groundwater? There appears to be a topographic barrier between the Tsagaan Nuur wetland and the Böön Tsagaan lake in this figure, but It is hard to appreciate if there is a topographic barrier n Figure 2. It would be helpful to plot the topographic barrier on Figure 2.

 

Response to Comment 40: We are afraid that illustrating the topographic barrier in Fig. 2 may increase the complexity of the figure. Instead of illustrating the topographic barrier, we have added the following sentence in the caption of Fig. 2 to clarify the location, as follows. We hope that this revision improves the manuscript.

“The altitudes of BTS-TN 10, 14, 16, and 18 are higher than those of the Böön Tsagaan lake and Tsagaan Nuur wetland (Supplementary Table 1), being a topographic barrier.”

 

 

Comment 41: Line 600: diageneted -> diagenetic.

 

Response to Comment 41: This has been modified.

 

 

Comment 42: Line 641 – 648: What would the source of water at the Tsagaan Nuur wetland be? Would it also be fed by the Baidrag river, or overland flow? Why would evaporation of water at the Tsagaan Nuur wetland lead to water so much more saline than at the lakes?

 

Response to Comment 42: The major water source is small transient rivers from the Khangai mountain area. Another water source is flooding of the Baidrag river in rainy seasons. The high concentration of salinity in the Tsagaan Nuur wetland is likely because of effective evaporation of surface water from shallow and widespread wetland. We have added an explanation of the water source of the Tsagaan Nuur wetland (see Response to Comment 39).

 

Reviewer 2 Report

The manuscript deals with a geochemical modelling of closed lakes in the Gobi Valley. The authors used data acquired during different field campaigns to constrain the role of groundwater to feed the lacustrine systems and then compare the obtained results to the Jezero and Gale landing sites of Mars.

Given the limited information we have on the past hydrological system on Mars, I’m fully supporting the search and characterisation of terrestrial analogues to better understand the martian settings, particularly the lacustrine systems, main goal for astrobiological investigation.
The proposed lakes in Mongolia may represents an interesting and compelling analogue.

A brief description of the geology of the study area is missing and should be added to provide information on the lithologies and tectonic setting which may strongly affect the circulation of the groundwater.

The authors compared the chemical/mineralogical composition of lacustrine water with recent and past sediments with the objective to better understand the chemical equilibrium of the water system.
However, the limited number of samples may introduce an error which has not been sufficiently discussed by the authors. Also, the sampling campaign have been carried out in different years which may introduce an additional uncertainty in the final results.
I’d appreciate a more complete discussion on the chemical differences between the rainy and dry season.

It’d be useful for the interpretation the semi-quantitative XRD measurements.

The text needs a better organisation and synthesis at places.

I suggest to rename of the sampling sites “stop XX” with the acronym of the geographic location to facilitate the reading of the figures.
Also some information (i.e. altitude and type of the sampling) that are provided in the supplementary materials should be included in the manuscript to support the interpretation of the overall amount of geochemical data.

For example Table 1 contains data acquired in two different campaigns (fall 2017 and summer 2018); I’d suggest to split it in two separate tables or indicate in consecutive rows the data acquired in different years for the same site.

The discussion is ripetitive at places and the reader may feel confused. I suggest to revise the text giving more emphasis to the main issues.

More comments are provided in the attached PDF.

Comments for author File: Comments.pdf

Author Response

Responses to comments from Reviewer #2 for ‘Hydrogeochemical study on closed-basin lakes in cold and semi-arid climates of the Valley of the Gobi Lakes, Mongolia: Implications for hydrology and water chemistry of paleolakes on Mars’ by Sekine, Y., et al.

 

We are grateful to reviewer #2 for the constructive comments and suggestions. Below, we give our responses in turn following each comment, with the reviewers’ comments being in Arial font and our responses being in Times font. 

Major Comment 1: A brief description of the geology of the study area is missing and should be added to provide information on the lithologies and tectonic setting which may strongly affect the circulation of the groundwater.

 

Response to Major Comment 1

To explain briefly the geological setting of the study area, we have added the following sentences in lines 123–131 of the revised manuscript:

“The Khangai mountains consist mainly of Precambrian and Paleozoic sedimentary rocks and Mesozoic granitic intrusions; whereas, the Altai Gobi mountains is formed by transpressional Cainozoic movements of Paleozoic volcanic and sedimentary rocks [33]. Major faults exist along with the Khangai and Altai Gobi mountains, intersecting the Baidrag and Tuin rivers [33]. The Valley of the Gobi Lakes is an elongated inter-montane depression between the Siberian craton and the Tarim craton [33]. The basins of the Valley of the Gobi Lakes are filled with Quaterary alluvial fans, lacustrine sediments, and aeolian deposits [28,33]. The alluvial, lacustrine, aeolian deposits are poorly cemented. The thickness of the alluvial, lacustrine, and aeolian deposits can reach 20–40 m or greater in the Valley of the Gobi Lakes, capable of storing groundwater in the subsurface.”

 

 

Major Comment 2: The authors compared the chemical/mineralogical composition of lacustrine water with recent and past sediments with the objective to better understand the chemical equilibrium of the water system.

However, the limited number of samples may introduce an error which has not been sufficiently discussed by the authors. Also, the sampling campaign have been carried out in different years which may introduce an additional uncertainty in the final results.

I’d appreciate a more complete discussion on the chemical differences between the rainy and dry season.

 

Response to Major Comment 2: We agree on the reviewer’s comment that the number and timing of the sampling are still insufficient to reconstruct the whole of groundwater hydrology in the Valley of the Gobi Lakes. Based on the comment, we have largely rewritten Sec. 5.1 by adding the following paragraphs to explain the uncertainties of our field surveys due to the limitation of the number of sampling sites and timing of survey.

“Although we performed regional-scale field surveys for the Valley of the Gobi Lakes, the sampling sites are still sparce compared to the whole of the catchment areas of the Bӧӧn Tsagaan and Orog lakes. In particular, there is only one sampling site in the west of the Bӧӧn Tsagaan lake (BTS west) and a few sampling sites around the Orog lake (Fig. 2). The water chemistry of BTS west exhibits a distinct chemical composition, which is characterized as the highest fraction of SO₄^2- in the anions among the sampling sites (Fig. 4). Thus, the groundwater seepage from the west of Bӧӧn Tsagaan lake, if occurred, could provide additional SO₄^2- to the lake, although SO₄^2--rich groundwater cannot be the major source of water to the lake to explain the chemistry of Bӧӧn Tsagaan lake (see Sec. 5.1.1 below). In addition, we performed field surveys in a rainy and dry season in different years (i.e., dry season in 2017 and rainy season in 2018). Thus, the seasonal variations in the water chemistry considered in the present study might partly include the annual variations of the water chemistry due to a change in precipitation.

To reduce the uncertainties caused by the limitation of the number of sampling sites and timing of survey, further investigations over the Valley of the Gobi Lakes in a single year would be required. Despite the uncertainties, we suggest that the following two conclusions of our field survey are robust. First is that the water chemistry of the lakes is distinct from those of the rivers and groundwaters. Second is that we find three types of water chemistry (i.e., groups A, B-east, and B-west) for the groundwaters and rivers despite the surveys in the different years. In this section, we discuss the interactions between the lake and groundwater (Sec. 5.1.1) and possible mechanisms to generate different types of groundwaters (Sec. 5.1.2).”

 

 

Major Comment 3: It’d be useful for the interpretation the semi-quantitative XRD measurements.

 

Response to Major Comment 3: Since there are large fractions of amorphous phase and/or organic matter in the lake sediment samples, (semi-)quantitative XRD analysis is not straightforward using our available analysis software. Instead, we have created a semi-quantitative table of major minerals found in the samples in new Supplementary Table 1. We hope that this revision could improve the manuscript.

 

 

Major Comment 4: The text needs a better organisation and synthesis at places. I suggest to rename of the sampling sites “stop XX” with the acronym of the geographic location to facilitate the reading of the figures.

 

Response to Major Comment 4: We have renamed the sampling site throughout the revised manuscript. In particular, we used “BTS” for the samples from near the Boon Tsagaan lake, “BTS-TN” for the samples from in between the Boon Tsagaan lake and Tsagaan Nuur wetland, and “TN” for the samples near the Tsagaan Nuur wetland. Based on these changes, we have modified Figs. 2, 3, 4, and 7, as well as Tables 2 and 3 of the revised manuscript.

 

 

Major Comment 5: Also some information (i.e. altitude and type of the sampling) that are provided in the supplementary materials should be included in the manuscript to support the interpretation of the overall amount of geochemical data.

 

Response to Major Comment 5: We have added new Table 1 that describes the locations and altitudes of the sampling sites in the main text.

 

 

Major Comment 6: For example Table 1 contains data acquired in two different campaigns (fall 2017 and summer 2018); I’d suggest to split it in two separate tables or indicate in consecutive rows the data acquired in different years for the same site.

 

Response to Major Comment 6: We have separated the original Table 1 into two, new Tables 2 (data of 2018’s survey) and 3 (data of 2017’s survey) in the revised manuscript.

 

 

Major Comment 7: The discussion is ripetitive at places and the reader may feel confused. I suggest to revise the text giving more emphasis to the main issues.

 

Response to Major Comment 7: We have deleted some repeated sentences in Discussion of the revised manuscript. In particular, we have largely rephrased Sec. 5.1. We hope that this could improve the readability.

 

 

Comment 8: More comments are provided in the attached PDF

 

Response to Comment 8: We appreciate the reviewer’s careful checks and comments. We have revised the manuscript according to the reviewer’s suggestions. As for more detailed responses, please see the attached PDF file.

 

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors have revised the manuscript taking into account the requested changes. I don't have further comment and recommend it for publication.