the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 30 Jun 2025
Crop salinization by intense pumping in regional discharge areas of an inland aquifer system (Cenozoic Duero basin aquifer, Spain)
Abstract. Salinization of crops irrigated with groundwaters in the Tordesillas area has been investigated to determine its cause. Hydrogeological, geophysical, and geochemical techniques reveal that regional saline groundwater flows through the Cenozoic aquifer system of the Duero Basin discharge into the Tordesillas area. Groundwater salinity increases below 150–200 depth. TDEM profiles indicate that salinity distribution is influenced by local and regional flow mixing, as well as by fault structures affecting the Cenozoic succession. Isotopic analyses (δ18O, δD, δ34S) suggest multiple sources of dissolved sulphate and evidence that regional groundwaters recharged at higher altitudes and/or lower temperatures.
Irrigation return flows do not noticeably contribute to salinization, as δ18O and δD data from boreholes in the Duero Floodplain do not show an evaporation trend. Instead, intensive groundwater pumping (from boreholes in the Duero River floodplain), particularly during the irrigation season, induces upwelling of saline groundwater. Piezometric records indicate that hydraulic potential at intermediate depths (about 100 m depth) decreases during pumping (summer), facilitating upwelling of deeper saline groundwaters. Salinity profiles confirm this process, demonstrating a shift from fresher to more saline conditions over time.
Groundwater management authorities must address this issue to prevent further salinization. These findings provide crucial insights for optimizing well design and identifying depths where groundwater is unsuitable for irrigation, ensuring sustainable water use in the region.
- Preprint
(4374 KB) - Metadata XML
-
Supplement
(152 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2025-2395', Anonymous Referee #1, 03 Jul 2025
-
AC1: 'Reply on RC1', Pedro Huerta, 26 Jul 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-2395/egusphere-2025-2395-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Pedro Huerta, 26 Jul 2025
-
RC2: 'Comment on egusphere-2025-2395', Anonymous Referee #2, 10 Oct 2025
The manuscript focuses on the crop salinization event that occurred in the Tordesillas region. Based on hydraulic head observations, TDEM geophysical surveys, chemical analyses, and stable isotope data, the authors proposed that summer irrigation pumping induced the upwelling of deep saline water, leading to salinization as the main cause. The research topic has certain significance for regional water resources management. However, the current manuscript lacks sufficient evidence in several key aspects (especially the quantitative relationship between pumping volume and salinity variation, geochemical end-member identification, and validation between TDEM and well data) to support the conclusion proposed by the authors. The paper requires substantial additional work (including more data, quantitative modeling, and end-member analysis) and major rewriting. Therefore, my recommendation is Reject. The main issues and revision suggestions are listed below.
1. The authors mainly infer that pumping-induced upwelling occurs based on the temporal correlation between the observed seasonal decline in water level and the simultaneous increase of EC in the middle layer (100–150 m). However, no detailed extraction records—such as pumping volume, well location, and pumping schedule—are provided. In addition, no numerical hydrodynamic or solute transport modeling is conducted to quantify the relationship between upwelling intensity and salinity response, which weakens the persuasiveness of the key conclusion.
It is recommended to supplement and present detailed pumping records of the main irrigation wells (including well locations, pumping volumes, pumping time series, borehole structure, and screen intervals). A simple hydrodynamic and solute transport simulation, or a simplified 1D/2D model, should be performed to test whether the observed water-level decline and EC variations can be reproduced under the recorded pumping intensity, thereby quantifying the upwelling flow rate, salt flux, and EC variation timescale.
2.The authors use δ¹⁸O/δD, δ³⁴S, and δ¹⁸O(SO₄) data to discuss water and sulfate sources, but they have not adequately sampled or characterized the potential “end-members” (e.g., local Paleogene/Miocene gypsum, agricultural sulfate fertilizers, river water, or near-surface return flow) in terms of isotopic and chemical composition. The conclusion that “irrigation return flow is insignificant” is based merely on the absence of an evaporation line in δ¹⁸O/δD, which is not sufficient.
It is suggested to supplement the analysis of potential source contributions, perform end-member mixing analysis (EMMA, SIAR, or a simple mass-balance approach) to quantitatively estimate the contributions of different end-members, and assess the associated uncertainties.
3. The authors conducted TDEM surveys and presented several geoelectric units (GU-1 to GU-4), but the calibration between resistivity and salinity, the discussion of inversion uncertainty, and the evaluation of model non-uniqueness are insufficient. Moreover, no validation was carried out using borehole lithology, downhole resistivity, or in-well conductivity data. As a result, the geophysical (TDEM) interpretation lacks strong quantitative support.
4. The conclusion emphasizes that “pumping-induced upwelling leads to salinization,” but no salt flux estimation (such as upward salt flux, irrigation-area salt input/output, or accumulation rate) is provided. It is recommended to estimate salt fluxes and perform simplified salt mass-balance calculations based on observed water levels, salinity, and pumping data, and to model long-term salinity evolution under different pumping strategies. This would enable quantitative management recommendations (e.g., allowable extraction thresholds or well-screening depth strategies).
5. The Results and Discussion sections overlap considerably, and the current section division is fragmented, resembling a technical report rather than a scientific article. The Introduction lacks the background and research progress on issues such as crop salinization induced by intensive pumping, and does not sufficiently review recent literature. Most references are outdated. It is recommended to add recent studies from the past five years concerning groundwater-extraction-induced salinization in inland basins, TDEM interpretation, and isotopic tracing, to strengthen the arguments and improve the logical focus of the paper.
Citation: https://doi.org/10.5194/egusphere-2025-2395-RC2 -
AC2: 'Reply on RC2', Pedro Huerta, 03 Nov 2025
- The authors mainly infer that pumping-induced upwelling occurs based on the temporal correlation between the observed seasonal decline in water level and the simultaneous increase of EC in the middle layer (100–150 m). However, no detailed extraction records—such as pumping volume, well location, and pumping schedule—are provided. In addition, no numerical hydrodynamic or solute transport modeling is conducted to quantify the relationship between upwelling intensity and salinity response, which weakens the persuasiveness of the key conclusion.
It is recommended to supplement and present detailed pumping records of the main irrigation wells (including well locations, pumping volumes, pumping time series, borehole structure, and screen intervals). A simple hydrodynamic and solute transport simulation, or a simplified 1D/2D model, should be performed to test whether the observed water-level decline and EC variations can be reproduced under the recorded pumping intensity, thereby quantifying the upwelling flow rate, salt flux, and EC variation timescale.
- Reply: In Section 3, we indicate that the annual groundwater extraction volume in the area is approximately 0.65 Hm³, with pumping mainly concentrated during the summer months, from May to September. Unfortunately, detailed pumping records are not available, as farmers usually do not report their pumping activities or the characteristics of their wells. However, interviews with farmers and well construction companies have allowed us to determine that the well screens are generally located between 80 and 150 m depth. The locations of the wells are shown in Fig. 2.
We agree with Reviewer 2 that a simple solute transport model could strengthen the conclusions of this manuscript. Nevertheless, the significant uncertainties regarding well data, pumping schedules, and discharge rates make it difficult to perform a reliable simulation.
- The authors use δ¹⁸O/δD, δ³⁴S, and δ¹⁸O(SO₄) data to discuss water and sulfate sources, but they have not adequately sampled or characterized the potential “end-members” (e.g., local Paleogene/Miocene gypsum, agricultural sulfate fertilizers, river water, or near-surface return flow) in terms of isotopic and chemical composition. The conclusion that “irrigation return flow is insignificant” is based merely on the absence of an evaporation line in δ¹⁸O/δD, which is not sufficient. It is suggested to supplement the analysis of potential source contributions, perform end-member mixing analysis (EMMA, SIAR, or a simple mass-balance approach) to quantitatively estimate the contributions of different end-members, and assess the associated uncertainties.
- Reply: We believe that the δ¹⁸O/δD composition of groundwater is adequately reported, including values measured for water from the Duero River (Table A2 and Fig. 8).
Regarding sulfate, only Triassic gypsum values have been included in Fig. 8B, whose δ³⁴S and δ¹⁸O(SO₄) values are consistent with those of dissolved sulfate in groundwater. Mesozoic gypsum occurs along the basin margins and is considered in the manuscript because our working hypothesis is that deep groundwater is recharged in these elevated areas.
Although limited sulfate isotope data exist for the Cenozoic basin infill (some previously published, others unpublished), there is no evidence of outcropping or subsurface gypsum deposits upstream of the deep groundwater flow paths, as indicated by piezometric data. Nevertheless, δ³⁴S values of Cenozoic gypsum (from the literature and our own measurements, unrelated to the study area) range widely from approximately +5‰ to +20‰, overlapping the δ³⁴S values of deep groundwater. However, their δ¹⁸O(SO₄) values are much higher (+15 to +25‰), such that the δ¹⁸O(SO₄)–δ³⁴S compositions of dissolved sulfate in deep wells and Cenozoic gypsum do not overlap (Fig. 8B).
δ¹⁸O(SO₄) and δ³⁴S values for the Duero River are reported in Table A2 and plotted in Fig. 8B (light blue filled circles). River sulfate plots at the lower end of the deep-well range. Irrigation return flows are difficult to isolate among shallow waters. Samples P-28 and P-30 (“dug wells” in Table A2; dark blue filled circles in Fig. 8B) likely represent the best available proxies for shallow waters potentially influenced by irrigation return flow. In Fig. 8B, these samples show the lightest isotopic compositions among the dataset but cannot account for the higher δ¹⁸O(SO₄) and δ³⁴S values observed in deep groundwater.
We do not have original isotope data for sulfate fertilizers; however, Vitoria et al. (2004; Environ. Sci. Technol., 38, 3254–3262) reported measurements for chemical fertilizers commercially available in Spain. We have no reason to believe that current products used in the study area differ significantly from those reported. According to their dataset, fertilizers classified as “straight” (i.e., with higher sulfate content) exhibit δ¹⁸O(SO₄) values near the lower end of our dissolved sulfate range, but δ³⁴S values that are distinctly lower—even lower than those measured in Duero River water. Therefore, fertilizers are considered an unlikely source of sulfate in deep groundwater.
Finally, average δ¹⁸O and δD values for deep groundwater in the Villafáfila area are plotted in Fig. 8A, and corresponding δ¹⁸O(SO₄) and δ³⁴S values are shown in Fig. 8B. Unlike the Tordesillas area, Villafáfila has little or no irrigated agriculture, so fertilizer use and irrigation return flow are minimal. Despite this, dissolved sulfate compositions in both areas are similar and match those of Mesozoic gypsum in the basin margins. This, together with the low δ¹⁸O and δD values of groundwater, supports our hypothesis that deep groundwater is recharged in the elevated basin margins and flows at depth toward the central areas.
- The authors conducted TDEM surveys and presented several geoelectric units (GU-1 to GU-4), but the calibration between resistivity and salinity, the discussion of inversion uncertainty, and the evaluation of model non-uniqueness are insufficient. Moreover, no validation was carried out using borehole lithology, downhole resistivity, or in-well conductivity data. As a result, the geophysical (TDEM) interpretation lacks strong quantitative support.
- Reply:
The geoelectric units identified in the TDEM profiles are consistent with the EC measurements obtained from wells. The high resistivities of GU-1 are confirmed by the low EC values measured in shallow (dug) wells. The progressive decrease in resistivity from GU-1 to GU-3 corresponds to the progressive increase in EC observed in well P42 (Figs. 4 and 5). As can be seen, both the change from GU-2 to GU-3 and the EC shift in P42 (June 2022) occur at approximately 500 m a.s.l. It is worth noting that TDEM sounding SEDT-4 is located at the same site as wells P42, P43, and P44, providing a direct comparison between datasets. We therefore consider that the validation of the TDEM interpretation is adequate.
The methodology section describes the procedure used to invert apparent resistivities to true resistivities. Lithological data from exploration boreholes (well logs and drill cuttings)—some of which are published in IGME (1980a, b)—indicate that sandstones and mudstones dominate the subsurface Cenozoic basin fill in this area. Variations in resistivity are thus interpreted to reflect differences in groundwater salinity. The geophysical interpretation is further supported by geological mapping and the regional geological framework, both of which are consistent with the inferred subsurface structure.
- The conclusion emphasizes that “pumping-induced upwelling leads to salinization,” but no salt flux estimation (such as upward salt flux, irrigation-area salt input/output, or accumulation rate) is provided. It is recommended to estimate salt fluxes and perform simplified salt mass-balance calculations based on observed water levels, salinity, and pumping data, and to model long-term salinity evolution under different pumping strategies. This would enable quantitative management recommendations (e.g., allowable extraction thresholds or well-screening depth strategies).
- Reply: Reviewer 2 suggests constructing a model to quantify salt fluxes. We agree that such an approach would be valuable; however, it represents a complex task that falls beyond the scope of the present manuscript and should be addressed in a separate study. In this work, our objective is to highlight the occurrence of a common issue in many non-marine aquifer systems and to propose a conceptual hypothesis supported by multiple lines of evidence.
The results of our research indicate that:
-
- The Tordesillas area functions as a groundwater discharge zone within the Cenozoic aquifer system, where groundwater flow exhibits an upward vertical component that discharges into the Duero River.
- The salinity distribution in the Cenozoic Duero aquifer, particularly in this area, reveals the presence of brackish groundwater below approximately 500 m a.s.l. (as shown by the EC profiles in well P42, the SEDT profiles in this paper, and in Huerta et al., 2021; Huerta et al., 2022; Nieto et al., 2020).
- Temporal monitoring of hydraulic head and EC in multiple piezometers shows the seasonal rise of the deep fresh–brackish water interface during the summer pumping period.
- The Results and Discussion sections overlap considerably, and the current section division is fragmented, resembling a technical report rather than a scientific article. The Introduction lacks the background and research progress on issues such as crop salinization induced by intensive pumping, and does not sufficiently review recent literature. Most references are outdated. It is recommended to add recent studies from the past five years concerning groundwater-extraction-induced salinization in inland basins, TDEM interpretation, and isotopic tracing, to strengthen the arguments and improve the logical focus of the paper.
- Reply: We consider that a scientific article should present the results clearly separated from their interpretation. The overlap between the Results and Discussion sections mentioned by Reviewer 2 arises from our deliberate effort to distinguish between data presentation and interpretation. In the Results section, we present the data obtained, while in the Discussion section we interpret these findings, compare them with similar processes and contexts, and evaluate alternative hypotheses.
The Introduction of the manuscript outlines the problem addressed, discusses the possible causes, and cites both classic and recent papers—many of which are review articles on the topic. Approximately 40% of the references cited in the Introduction are from the past five years. Nevertheless, we acknowledge the reviewer’s suggestion and will incorporate a more comprehensive review of recent literature related to groundwater-extraction-induced salinization, TDEM applications, and isotopic tracing to further strengthen the scientific context and focus of the paper.
Citation: https://doi.org/10.5194/egusphere-2025-2395-AC2
-
AC2: 'Reply on RC2', Pedro Huerta, 03 Nov 2025
Status: closed
-
RC1: 'Comment on egusphere-2025-2395', Anonymous Referee #1, 03 Jul 2025
Unfortunately, this paper is not suitable for publication in an international journal such as HESS. It has no novel aspects and would not be of interest to the broader academic community. The conclusions are reasonable, but the main finding that groundwater pumping causes changes to aquifer salinities is not unexpected and not novel. In its present form it is more suited to a regional journal. Case studies are certainly acceptable in the international literature, but they need to add to our general understanding; so, indications of what new ideas come out of a study and what researchers working on similar projects elsewhere can take from it are needed.
The limited and parochial scope of the paper is evident in the Introduction, which is focused on the local issue of crop yields being impacted by salinization. Salinization of water resources is a major problem globally, but there is no effort here to review the global understanding or to put this study into a broader context. The Aims are also very specific to understanding the local hydrogeology and the Conclusions are just a restatement of the specific findings of the research with no indication of how or why this research is of general interest. Even within the local context, the end of the paper is underwhelming with a general suggestion that these data should help management (without specifying how).
There are several issues with the data and its interpretation.
- Data limitations. The geochemical interpretations are based on a standard set of parameters (major ions, water stable isotopes, sulfate stable isotopes) from a small number of groundwater samples (12 in total). This is a very limited dataset that is unlikely to yield much insight into processes. The groundwater samples are also poorly characterized – the text and Table A1 lists samples as being <40 m and >40 m, but what are the exact depths? It is much more straightforward to interpret geochemistry data from wells that have short screens and which sample water from a specific aquifer than from wells with long screens that integrate water from several layers. Section 5.2.1 divides the sequence up into several GU’s and it would be good to be able to link the geochemical data with those units, which is not possible with the current reporting.
- Major ions. The interpretation of the major ion geochemistry does not really add much. Section 5.3 is descriptive and mainly defines groundwater types and their distribution. Section 6 mainly makes use of the overall groundwater TDS and salinity but not really the major ion geochemistry. Mixing is discussed in Section 6.2, but there is little attempt to quantify it or justify the conclusions. Even if quantification is not possible, identifying the end-members and showing the mixing on the Piper diagram or other plots would help. However, it is difficult to use small datasets to produce robust conclusions about processes such as mixing, which makes this discussion speculative. Some of the changes in water chemistry may reflect processes such as mineral dissolution and precipitation, which can produce systematic changes in water chemistry with salinity (e.g., Herczeg et al., 2001. Origin of dissolved salts in a large, semi-arid groundwater system: Murray Basin, Australia. Marine and Freshwater Resources, 52, 41-52, https://doi.org/ 10.1071/MF00040); again, this needs consideration.
- Water stable isotopes. The comparison of the groundwater stable isotopes with those of the weighted mean precipitation (Section 5.4) implicitly assumes that the groundwater is recharged by precipitation with that isotopic composition and any deviations occurs due to recharge at elevation (Section 6.4). However, it is likely that recharge occurs preferentially from high rainfall events or during winter when evapotranspiration rates are low and the comparison needs to consider this. Is it possible that the isotopic composition of the rainfall that causes recharge is different from the annual mean – for example, recharge mainly from winter rainfall, which commonly has lower stable isotope values than the mean, may explain the observations.
- Water stable isotopes. The observation that the stable isotopes do not define an evaporation trend may not rule out irrigation returns. Open-system evaporation in surface water bodies (pools, lakes etc) does produce distinctive isotopic trends. However, transpiration does not and it is not clear whether evaporation from within the soils where the relative humidity is higher fractionates stable isotopes to a large degree. There are plenty of examples of saline groundwater caused by evapotranspiration where the stable isotopes lie close to the MWL. This would include the deep groundwater in this region – looking at the major ion geochemistry and description of the aquifers, the high salinities are probably the result of evapotranspiration (there are no evaporites reported and halite dissolution produces a distinct NaCl-type geochemistry that is different to what is shown in Fig. 6). Yet the stable isotopes lie very close to the MWL. Again, the interpretation of data needs to be better justified.
- Sulfur isotopes. The discussion of the sulfur stable isotopes (section 6.4) is also general and not well justified. The isotopic values are interpreted as solely representing gypsum dissolution in the aquifers without consideration of whether other sources of sulfur (e.g., pyrite) might be present or whether fractionation due to processes such as bacterial sulfate reduction (which is common in groundwater globally) may have occurred. The conclusion that the Tordesillas groundwaters have sulfate derived from several sources is untested (is that consistent with the other data and the hydrology?). Similar comments apply to the conclusion that the isotopes show mixing in the river. As with the other datasets, you need to justify potential interesting conclusions such as this rather than just making assertions.
- Integration with the geophysics data. Partially due to the lack of detail regarding sample depth, it is difficult to link the geochemistry with the geophysics data. The geophysics results (discussed in Section 6.1) are presented separately to the geochemistry. Integrating both halves of the work would help the study.
I have not gone through the paper in more detail, as I cannot see that the data can be woven into a story that is of sufficient interest for this journal. It is never pleasant to receive negative reviews, but I would encourage the authors to see if they can bring more rigor to the study and consider how generally interesting / novel this work is, which will dictate where it should be submitted.
Citation: https://doi.org/10.5194/egusphere-2025-2395-RC1 -
AC1: 'Reply on RC1', Pedro Huerta, 26 Jul 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-2395/egusphere-2025-2395-AC1-supplement.pdf
-
RC2: 'Comment on egusphere-2025-2395', Anonymous Referee #2, 10 Oct 2025
The manuscript focuses on the crop salinization event that occurred in the Tordesillas region. Based on hydraulic head observations, TDEM geophysical surveys, chemical analyses, and stable isotope data, the authors proposed that summer irrigation pumping induced the upwelling of deep saline water, leading to salinization as the main cause. The research topic has certain significance for regional water resources management. However, the current manuscript lacks sufficient evidence in several key aspects (especially the quantitative relationship between pumping volume and salinity variation, geochemical end-member identification, and validation between TDEM and well data) to support the conclusion proposed by the authors. The paper requires substantial additional work (including more data, quantitative modeling, and end-member analysis) and major rewriting. Therefore, my recommendation is Reject. The main issues and revision suggestions are listed below.
1. The authors mainly infer that pumping-induced upwelling occurs based on the temporal correlation between the observed seasonal decline in water level and the simultaneous increase of EC in the middle layer (100–150 m). However, no detailed extraction records—such as pumping volume, well location, and pumping schedule—are provided. In addition, no numerical hydrodynamic or solute transport modeling is conducted to quantify the relationship between upwelling intensity and salinity response, which weakens the persuasiveness of the key conclusion.
It is recommended to supplement and present detailed pumping records of the main irrigation wells (including well locations, pumping volumes, pumping time series, borehole structure, and screen intervals). A simple hydrodynamic and solute transport simulation, or a simplified 1D/2D model, should be performed to test whether the observed water-level decline and EC variations can be reproduced under the recorded pumping intensity, thereby quantifying the upwelling flow rate, salt flux, and EC variation timescale.
2.The authors use δ¹⁸O/δD, δ³⁴S, and δ¹⁸O(SO₄) data to discuss water and sulfate sources, but they have not adequately sampled or characterized the potential “end-members” (e.g., local Paleogene/Miocene gypsum, agricultural sulfate fertilizers, river water, or near-surface return flow) in terms of isotopic and chemical composition. The conclusion that “irrigation return flow is insignificant” is based merely on the absence of an evaporation line in δ¹⁸O/δD, which is not sufficient.
It is suggested to supplement the analysis of potential source contributions, perform end-member mixing analysis (EMMA, SIAR, or a simple mass-balance approach) to quantitatively estimate the contributions of different end-members, and assess the associated uncertainties.
3. The authors conducted TDEM surveys and presented several geoelectric units (GU-1 to GU-4), but the calibration between resistivity and salinity, the discussion of inversion uncertainty, and the evaluation of model non-uniqueness are insufficient. Moreover, no validation was carried out using borehole lithology, downhole resistivity, or in-well conductivity data. As a result, the geophysical (TDEM) interpretation lacks strong quantitative support.
4. The conclusion emphasizes that “pumping-induced upwelling leads to salinization,” but no salt flux estimation (such as upward salt flux, irrigation-area salt input/output, or accumulation rate) is provided. It is recommended to estimate salt fluxes and perform simplified salt mass-balance calculations based on observed water levels, salinity, and pumping data, and to model long-term salinity evolution under different pumping strategies. This would enable quantitative management recommendations (e.g., allowable extraction thresholds or well-screening depth strategies).
5. The Results and Discussion sections overlap considerably, and the current section division is fragmented, resembling a technical report rather than a scientific article. The Introduction lacks the background and research progress on issues such as crop salinization induced by intensive pumping, and does not sufficiently review recent literature. Most references are outdated. It is recommended to add recent studies from the past five years concerning groundwater-extraction-induced salinization in inland basins, TDEM interpretation, and isotopic tracing, to strengthen the arguments and improve the logical focus of the paper.
Citation: https://doi.org/10.5194/egusphere-2025-2395-RC2 -
AC2: 'Reply on RC2', Pedro Huerta, 03 Nov 2025
- The authors mainly infer that pumping-induced upwelling occurs based on the temporal correlation between the observed seasonal decline in water level and the simultaneous increase of EC in the middle layer (100–150 m). However, no detailed extraction records—such as pumping volume, well location, and pumping schedule—are provided. In addition, no numerical hydrodynamic or solute transport modeling is conducted to quantify the relationship between upwelling intensity and salinity response, which weakens the persuasiveness of the key conclusion.
It is recommended to supplement and present detailed pumping records of the main irrigation wells (including well locations, pumping volumes, pumping time series, borehole structure, and screen intervals). A simple hydrodynamic and solute transport simulation, or a simplified 1D/2D model, should be performed to test whether the observed water-level decline and EC variations can be reproduced under the recorded pumping intensity, thereby quantifying the upwelling flow rate, salt flux, and EC variation timescale.
- Reply: In Section 3, we indicate that the annual groundwater extraction volume in the area is approximately 0.65 Hm³, with pumping mainly concentrated during the summer months, from May to September. Unfortunately, detailed pumping records are not available, as farmers usually do not report their pumping activities or the characteristics of their wells. However, interviews with farmers and well construction companies have allowed us to determine that the well screens are generally located between 80 and 150 m depth. The locations of the wells are shown in Fig. 2.
We agree with Reviewer 2 that a simple solute transport model could strengthen the conclusions of this manuscript. Nevertheless, the significant uncertainties regarding well data, pumping schedules, and discharge rates make it difficult to perform a reliable simulation.
- The authors use δ¹⁸O/δD, δ³⁴S, and δ¹⁸O(SO₄) data to discuss water and sulfate sources, but they have not adequately sampled or characterized the potential “end-members” (e.g., local Paleogene/Miocene gypsum, agricultural sulfate fertilizers, river water, or near-surface return flow) in terms of isotopic and chemical composition. The conclusion that “irrigation return flow is insignificant” is based merely on the absence of an evaporation line in δ¹⁸O/δD, which is not sufficient. It is suggested to supplement the analysis of potential source contributions, perform end-member mixing analysis (EMMA, SIAR, or a simple mass-balance approach) to quantitatively estimate the contributions of different end-members, and assess the associated uncertainties.
- Reply: We believe that the δ¹⁸O/δD composition of groundwater is adequately reported, including values measured for water from the Duero River (Table A2 and Fig. 8).
Regarding sulfate, only Triassic gypsum values have been included in Fig. 8B, whose δ³⁴S and δ¹⁸O(SO₄) values are consistent with those of dissolved sulfate in groundwater. Mesozoic gypsum occurs along the basin margins and is considered in the manuscript because our working hypothesis is that deep groundwater is recharged in these elevated areas.
Although limited sulfate isotope data exist for the Cenozoic basin infill (some previously published, others unpublished), there is no evidence of outcropping or subsurface gypsum deposits upstream of the deep groundwater flow paths, as indicated by piezometric data. Nevertheless, δ³⁴S values of Cenozoic gypsum (from the literature and our own measurements, unrelated to the study area) range widely from approximately +5‰ to +20‰, overlapping the δ³⁴S values of deep groundwater. However, their δ¹⁸O(SO₄) values are much higher (+15 to +25‰), such that the δ¹⁸O(SO₄)–δ³⁴S compositions of dissolved sulfate in deep wells and Cenozoic gypsum do not overlap (Fig. 8B).
δ¹⁸O(SO₄) and δ³⁴S values for the Duero River are reported in Table A2 and plotted in Fig. 8B (light blue filled circles). River sulfate plots at the lower end of the deep-well range. Irrigation return flows are difficult to isolate among shallow waters. Samples P-28 and P-30 (“dug wells” in Table A2; dark blue filled circles in Fig. 8B) likely represent the best available proxies for shallow waters potentially influenced by irrigation return flow. In Fig. 8B, these samples show the lightest isotopic compositions among the dataset but cannot account for the higher δ¹⁸O(SO₄) and δ³⁴S values observed in deep groundwater.
We do not have original isotope data for sulfate fertilizers; however, Vitoria et al. (2004; Environ. Sci. Technol., 38, 3254–3262) reported measurements for chemical fertilizers commercially available in Spain. We have no reason to believe that current products used in the study area differ significantly from those reported. According to their dataset, fertilizers classified as “straight” (i.e., with higher sulfate content) exhibit δ¹⁸O(SO₄) values near the lower end of our dissolved sulfate range, but δ³⁴S values that are distinctly lower—even lower than those measured in Duero River water. Therefore, fertilizers are considered an unlikely source of sulfate in deep groundwater.
Finally, average δ¹⁸O and δD values for deep groundwater in the Villafáfila area are plotted in Fig. 8A, and corresponding δ¹⁸O(SO₄) and δ³⁴S values are shown in Fig. 8B. Unlike the Tordesillas area, Villafáfila has little or no irrigated agriculture, so fertilizer use and irrigation return flow are minimal. Despite this, dissolved sulfate compositions in both areas are similar and match those of Mesozoic gypsum in the basin margins. This, together with the low δ¹⁸O and δD values of groundwater, supports our hypothesis that deep groundwater is recharged in the elevated basin margins and flows at depth toward the central areas.
- The authors conducted TDEM surveys and presented several geoelectric units (GU-1 to GU-4), but the calibration between resistivity and salinity, the discussion of inversion uncertainty, and the evaluation of model non-uniqueness are insufficient. Moreover, no validation was carried out using borehole lithology, downhole resistivity, or in-well conductivity data. As a result, the geophysical (TDEM) interpretation lacks strong quantitative support.
- Reply:
The geoelectric units identified in the TDEM profiles are consistent with the EC measurements obtained from wells. The high resistivities of GU-1 are confirmed by the low EC values measured in shallow (dug) wells. The progressive decrease in resistivity from GU-1 to GU-3 corresponds to the progressive increase in EC observed in well P42 (Figs. 4 and 5). As can be seen, both the change from GU-2 to GU-3 and the EC shift in P42 (June 2022) occur at approximately 500 m a.s.l. It is worth noting that TDEM sounding SEDT-4 is located at the same site as wells P42, P43, and P44, providing a direct comparison between datasets. We therefore consider that the validation of the TDEM interpretation is adequate.
The methodology section describes the procedure used to invert apparent resistivities to true resistivities. Lithological data from exploration boreholes (well logs and drill cuttings)—some of which are published in IGME (1980a, b)—indicate that sandstones and mudstones dominate the subsurface Cenozoic basin fill in this area. Variations in resistivity are thus interpreted to reflect differences in groundwater salinity. The geophysical interpretation is further supported by geological mapping and the regional geological framework, both of which are consistent with the inferred subsurface structure.
- The conclusion emphasizes that “pumping-induced upwelling leads to salinization,” but no salt flux estimation (such as upward salt flux, irrigation-area salt input/output, or accumulation rate) is provided. It is recommended to estimate salt fluxes and perform simplified salt mass-balance calculations based on observed water levels, salinity, and pumping data, and to model long-term salinity evolution under different pumping strategies. This would enable quantitative management recommendations (e.g., allowable extraction thresholds or well-screening depth strategies).
- Reply: Reviewer 2 suggests constructing a model to quantify salt fluxes. We agree that such an approach would be valuable; however, it represents a complex task that falls beyond the scope of the present manuscript and should be addressed in a separate study. In this work, our objective is to highlight the occurrence of a common issue in many non-marine aquifer systems and to propose a conceptual hypothesis supported by multiple lines of evidence.
The results of our research indicate that:
-
- The Tordesillas area functions as a groundwater discharge zone within the Cenozoic aquifer system, where groundwater flow exhibits an upward vertical component that discharges into the Duero River.
- The salinity distribution in the Cenozoic Duero aquifer, particularly in this area, reveals the presence of brackish groundwater below approximately 500 m a.s.l. (as shown by the EC profiles in well P42, the SEDT profiles in this paper, and in Huerta et al., 2021; Huerta et al., 2022; Nieto et al., 2020).
- Temporal monitoring of hydraulic head and EC in multiple piezometers shows the seasonal rise of the deep fresh–brackish water interface during the summer pumping period.
- The Results and Discussion sections overlap considerably, and the current section division is fragmented, resembling a technical report rather than a scientific article. The Introduction lacks the background and research progress on issues such as crop salinization induced by intensive pumping, and does not sufficiently review recent literature. Most references are outdated. It is recommended to add recent studies from the past five years concerning groundwater-extraction-induced salinization in inland basins, TDEM interpretation, and isotopic tracing, to strengthen the arguments and improve the logical focus of the paper.
- Reply: We consider that a scientific article should present the results clearly separated from their interpretation. The overlap between the Results and Discussion sections mentioned by Reviewer 2 arises from our deliberate effort to distinguish between data presentation and interpretation. In the Results section, we present the data obtained, while in the Discussion section we interpret these findings, compare them with similar processes and contexts, and evaluate alternative hypotheses.
The Introduction of the manuscript outlines the problem addressed, discusses the possible causes, and cites both classic and recent papers—many of which are review articles on the topic. Approximately 40% of the references cited in the Introduction are from the past five years. Nevertheless, we acknowledge the reviewer’s suggestion and will incorporate a more comprehensive review of recent literature related to groundwater-extraction-induced salinization, TDEM applications, and isotopic tracing to further strengthen the scientific context and focus of the paper.
Citation: https://doi.org/10.5194/egusphere-2025-2395-AC2
-
AC2: 'Reply on RC2', Pedro Huerta, 03 Nov 2025
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 739 | 49 | 23 | 811 | 29 | 14 | 26 |
- HTML: 739
- PDF: 49
- XML: 23
- Total: 811
- Supplement: 29
- BibTeX: 14
- EndNote: 26
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Unfortunately, this paper is not suitable for publication in an international journal such as HESS. It has no novel aspects and would not be of interest to the broader academic community. The conclusions are reasonable, but the main finding that groundwater pumping causes changes to aquifer salinities is not unexpected and not novel. In its present form it is more suited to a regional journal. Case studies are certainly acceptable in the international literature, but they need to add to our general understanding; so, indications of what new ideas come out of a study and what researchers working on similar projects elsewhere can take from it are needed.
The limited and parochial scope of the paper is evident in the Introduction, which is focused on the local issue of crop yields being impacted by salinization. Salinization of water resources is a major problem globally, but there is no effort here to review the global understanding or to put this study into a broader context. The Aims are also very specific to understanding the local hydrogeology and the Conclusions are just a restatement of the specific findings of the research with no indication of how or why this research is of general interest. Even within the local context, the end of the paper is underwhelming with a general suggestion that these data should help management (without specifying how).
There are several issues with the data and its interpretation.
I have not gone through the paper in more detail, as I cannot see that the data can be woven into a story that is of sufficient interest for this journal. It is never pleasant to receive negative reviews, but I would encourage the authors to see if they can bring more rigor to the study and consider how generally interesting / novel this work is, which will dictate where it should be submitted.