Community Comment (CC1):

General comments:

Very good and novel research in the area of deep hydrogeology with a variety of applications in the geo-energy sector. However, some detail is missing. Please, consider the following minor comments to improve the manuscript before publication.

Response:

We sincerely thank the community member for the encouraging and constructive feedback. We appreciate the recognition of the novelty and significance of our research in deep hydrogeology and its relevance to geo-energy applications.

We agree that the manuscript can benefit from additional detail, and we have carefully addressed all the minor comments and suggestions to enhance the clarity and completeness of the work. We believe the revised version better reflects the scope and contributions of our study.

 As per the journal’s submission guidelines, we are first submitting our detailed responses to the reviewer’s comments. Following this, we will submit the revised manuscript reflecting all the suggested changes.

Specific comments:

Comment 1:

Lines 69-72. “Consideration of hydraulic properties is crucial in groundwater evaluations. Permeability is the most popular aquifer measure and is mainly used to assess the water-holding capacity of rocks all over the world”. Insert these papers where there is discussion on the role of geophysical and hydrogeological methods to detect the hydraulic properties of fractured rocks to inform flow models in granites, metamorphic and sandstone lithologies.

- Medici, G., Ling, F., Shang, J. 2023. Review of discrete fracture network characterization for geothermal energy extraction. Frontiers in Earth Science 11, 1328397.

- McKeown, C., Haszeldine, R.S., Couples, G.D. 1999. Mathematical modelling of groundwater flow at Sellafield, UK. Engineering Geology 52(3-4), 231-250.

Response 1:

The following revision was made to improve clarity and integrate the suggested references:

“Consideration of hydraulic properties is crucial in groundwater evaluations. Permeability is one of the most widely used aquifer parameters for assessing the water-holding and transmitting capacity of rocks across the globe. In fractured rock environments, such as granites, metamorphic, and sandstone formations, fluid flow is primarily governed by the geometry and connectivity of fractures rather than the rock matrix itself. Therefore, accurately characterizing hydraulic properties in these settings requires integrated approaches. Recent studies emphasize the role of combining geophysical and hydrogeological methods to detect and model these hydraulic properties effectively (McKeown et al., 1999; Medici et al., 2023). These approaches are essential for improving the reliability of flow models and for guiding groundwater management and geo-energy extraction strategies in complex geological settings.”

Comment 2:

Lines 146-152. Lots of multiple objectives (5). Please, clarify the general aim of your hydrogeological research.

Response 2:

ORIGINAL:

“The primary goals of this study were as follows: (1) to rapidly predict two- and three dimensional k models using geophysical methods; (2) to reliably assess the hydrogeological properties of rock formations for deep groundwater assessments in challenging geological settings; (3) to minimize costly boreholes and maximize the use of scarce drilling resources to collect hydrogeological data over large areas; (4) to decrease uncertainties in hydrogeological models; and (5) to promote the use of non-invasive geophysical techniques for hard rock groundwater investigations instead of costly drilling that can damage the rock.”

REVISED:

"The primary aim of this study is to develop and implement a geophysical-based approach for accurately predicting the spatial distribution of permeability (k) in deep, hard rock environments. By integrating CSAMT data with strategically selected borehole measurements, this research enhances the two and three dimensional assessment of hydrogeological properties across various rock types in geologically complex settings, reduces reliance on extensive and costly drilling, and promotes the use of non-invasive geophysical techniques for deep groundwater exploration."

Comment 3:

Lines 173-181. The geometrical relation between the different lithologies is unclear.

Response 3:

ORIGINAL:

“Intruding rocks from the Indosinian, Caledonian, and Yanshanian eras are among the many geological formations and periods represented in the study region. Other layers from the Paleogene period are also present. The most common types of rock that have been discovered are sandstone, granite, and hornstone. The complex Kaiping concave fault and fold systems were the dominant geological features in the project region, which were developed as a result of magmatic processes and various structures (Qin, 2017). Emergence of joint fissured features symbolizes the various tectono-geological periods, with the local tectonic line corresponding with the faults strike, especially in the northeast orientation (Yang et al., 2021)”.

REVISED:

“The study area exhibits a complex and diverse geological history, characterized by well-defined geometrical relationships among various lithologies. These formations and structural features are the result of multiple tectono-magmatic events spanning several geological periods. Intrusive rocks from the Indosinian (Late Triassic), Caledonian (Silurian–Devonian), and Yanshanian (Jurassic–Cretaceous) orogenies are well-represented, indicating a long sequence of crustal deformation and magmatic activity. These intrusions are primarily composed of granitic bodies, which suggest deep-seated magmatic processes associated with continental collision and subduction zones. In addition to these intrusive phases, sedimentary strata from the Paleogene period are also present, reflecting a later stage of basin development with fluvial and lacustrine depositional environments. Among the most prevalent rock types encountered in the region are sandstone, granite, and hornstone. Sandstone reflects high-energy sedimentary deposition. Granite indicates deep magmatic intrusions likely associated with Yanshanian tectonics. Hornstone (hornfels) results from contact metamorphism caused by magma intruding sedimentary rocks. The structural framework of the region is dominated by the Kaiping concave fault and fold system, a geologically significant and highly deformed zone that reflects multiple deformation episodes (Qin, 2017). These structures were primarily shaped by magmatic intrusions, crustal movements, and regional stress regimes. The presence of extensive jointed and fissured zones throughout the rock mass further supports a history of dynamic tectonic activity. These joints often serve as secondary permeability pathways and are critical in controlling groundwater flow in the fractured rock environment. Importantly, the orientation of these structural features, including faults and joints, is often aligned with northeast-trending tectonic lines, which are consistent with broader regional stress directions (Yang et al., 2021). This relationship among lithologies and structural features plays a critical role in controlling groundwater flow and permeability distribution.”

Comment 4:

Lines 173-181. The detail is not enough on presence of faults. Which type of faults?

Response 4

More details on the presence of faults:

“The structural analysis of the study area reveals a combination of fault types influenced by multiple tectonic phases. The presence of fold systems indicates compressional tectonics, primarily associated with reverse and thrust faults, likely developed during orogenic events such as the Caledonian and Indosinian periods. Additionally, the dominant northeast-oriented fault strikes, which align with broader regional tectonic trends, suggest a strong component of strike-slip movement. These strike-slip faults are typically linked to late-stage tectonic adjustments, particularly during the Yanshanian orogenic phase, and often coexist with complex fault-fold geometries, further complicating the subsurface structure.”

Comment 5:

Lines 173-181. Nature of the joints? I am talking about the tectonic genesis.

Response 5:

Explained as:

“The joint fissure systems observed within the sandstone, granite, and hornstone units are predominantly of tectonic origin, representing brittle deformation features formed in response to regional stress regimes associated with multiple orogenic and magmatic events. These joints reflect the structural imprint of successive tectonic episodes, particularly the Caledonian, Indosinian, and Yanshanian orogenies. Systematic joint orientations, especially those aligned with the dominant northeast-trending fault systems, indicate their genetic link to regional tectonic stress fields. The geometry, spacing, and persistence of these joints vary with lithology and are closely tied to the complex tectonic evolution and structural framework governed by the Kaiping fold-and-fault system.”

Comment 6:

Line 538. I prefer “Discussion”. You have a unique discussion on a scientific paper where you face different topics. This point also depends on the guidelines.

Response 6:

Thank you for your valuable suggestion. We agree that “Discussion” is a more appropriate and conventional title for this section, as it aligns with standard scientific writing practices for presenting and interpreting key findings. Accordingly, we have changed the section title from “Discussions” to “Discussion” in compliance with the journal’s formatting guidelines. Furthermore, the entire Discussion section has been thoroughly revised and enhanced based on the suggestions provided by both the reviewers and the community, with the aim of improving clarity, depth, and overall scientific value.

Comment 7:

Lines 600-837. Insert the relevant literature suggested above on the hydraulic properties of deep aquifers in a variety of sites worldwide.

Response 7:

The relevant literature suggested above has been incorporated into the revised References section of the manuscript.

Comment 8:

Figure 1. Letters are too small in both the figures. Please, make the figure larger.

Response 8:

Fig. 1 has been improved. Please see the revised figures at the end of the response/comments section in the attached file.

Comment 9:

Figure 1. Pay lot of attention of figure 1b. This is a conceptual model and you can get citations from the figure. Make the figure larger and increase the font of the words.

Response 9

Fig. 1b has been improved accordingly. Please see the revised figures at the end of the response/comments section in the attached file.

Comment 10:

Figure 2. There is room to make the figure larger.

Response 10:

Fig 2 has been enlarged. Please see the revised figures at the end of the response/comments section in the attached file.

Comment 11:

Figure 4. Check the depth of the boreholes.

Response 11:

The depth of boreholes in the updated Fig. 5-7 has been corrected. Please see the revised figures at the end of the response/comments section in the attached file.

Comment 12:

Figure 9. The words are too small. The figure is difficult to read. Please, improve it.

Response 12:

The updated Fig.12 has been improved. Please see the revised figures at the end of the response/comments section in the attached file.

Community Comment (CC2):

The authors submitted the same manuscript to JoH.

Response:

We appreciate the comment. The manuscript submitted to the Journal of Hydrology (JoH) focused on a different hydraulic parameter, hydraulic conductivity, whereas the current submission emphasizes permeability estimation using CSAMT data. To avoid any overlap and maintain academic integrity, we have promptly withdrawn the JoH submission in response to this comment.

Referee Comments (RC2):

This paper addresses the challenges associated with current groundwater exploration and evaluates the advantages and disadvantages of various methods for measuring hydraulic parameters. The author highlights the application of a novel approach, by using Controlled Source Audio-Frequency Magnetotellurics (CSAMT) method, which is employed to estimate 2D and 3D permeability at depths exceeding 1 km in highly heterogeneous rock environments. The study presents its methods and findings in a well-structured manner, offering insights into deep groundwater exploration.

However, certain assertions appear overly generalized and could benefit from further substantiation. Additionally, more detailed descriptions of the methodologies and the study area would enhance the clarity, reproducibility, and robustness of the research.

Response:

We sincerely thank the anonymous reviewer for their insightful and constructive feedback, which has significantly contributed to enhancing the quality of our work. We have made every effort to revise the manuscript thoroughly in line with the reviewer’s suggestions.

In the revised version, we have expanded the descriptions of both the study area and the methodologies to provide greater clarity and context. Additional explanations and supporting information have also been included to substantiate our findings and address the points raised.

As per the journal’s submission guidelines, we are first submitting our detailed responses to the reviewer’s comments. Following this, we will submit the revised manuscript reflecting all the suggested changes.

Specific comments:

Comment 1:

Line108-119: the author might consider adding a little more evidence of the reason that CSAMT is selected for this study. For example, the author stated that VES method is used to evaluate groundwater resources in a single dimension by a broad of previous studies, but did not illustrate the background about why they did not use other methods, like CSAMT or ERT. Additionally, the author states that there are three main methods, but there are very few examples or introductions about ERT in this paragraph.

Response 1:

The entire paragraph from Lines 88–119 has been revised for improved clarity and structure in the updated manuscript. Both the original and the revised versions of the paragraph are provided below for comparison.

ORIGINAL PARAGRAPH:

“A number of prior groundwater investigations have made use of geophysical techniques (Bentley and Gharibi, 2004; Yadav and Singh, 2007; Fu et al., 2013; Vouillamoz et al., 2014; Robinson et al., 2016; Lin et al., 2018; Kouadio et al., 2020; Abbas et al., 2022; Kouadio et al., 2023; Zhang et al., 2024). A number of studies have shown that geophysical procedures outperform drilling techniques in terms of speed, ease of use, cost, and lack of invasiveness (Hu et al., 2013; Lin et al., 2018; Di et al., 2020; Fusheng et al., 2022; Hasan et al., 2024). Additionally, they are capable of conducting thorough geological evaluations in both the vertical and horizontal planes (Fu et al., 2013; Hasan et al., 2021). These methods are superior to others when it comes to collecting hydrogeological data from various subterranean habitats (Niwas and De Lima, 2003; Wynn et al., 2016; Kouadio et al., 2023). Groundwater studies nowadays often include resistivity surveys. Resistivity methods offer a broader resistivity range compared to other geophysical parameters, which is a major advantage (Bentley and Gharibi, 2004; Camporese et al., 2011; Robinson et al., 2016). The three main methods for measuring resistivity are the controlled source audio-frequency magnetotellurics (CSAMT), vertical electoral soundings (VES), and electrical resistivity tomography (ERT) (Soupios et al., 2007; Di et al., 2020; Zhang et al., 2024). Niwas and De Lima (2003), Soupios et al. (2007), Majumdar and Das (2011), Nwosu et al. (2013), Hasan et al. (2021), and Asfahani (2023) are among the previous groundwater-based geophysical studies that primarily utilized the VES method to evaluate groundwater resources in a single dimension. It is unusual to evaluate aquifer yield at great depths in hard rock terrains using two- and three-dimensional hydraulic properties. Recent studies have demonstrated that CSAMT, which aims to gather extensive subsurface data at very deep depths using 2D/3D evaluations, is the most cost-effective and appropriate geophysical method for researching hard rock (Smith and Booker, 1991; Simpson and Bahr, 2005; Bai et al., 2010; Fu et al., 2013; Hu et al., 2013; Wang et al., 2015; Wynn et al., 2016; Di et al., 2020; Zhang et al., 2021; Kouadio et al., 2023; Hasan et al., 2024). Advantages of CSAMT over other geophysical research methods include its lower cost, its responsiveness to low-resistance rocks, and its ease of usage in challenging topographic circumstances (An et al., 2016; Kouadio et al., 5 115 2020; Zhang et al., 2021). Compared to most geophysical technologies, including ERT, CSAMT's subsurface assessment capabilities are superior due to its depth capacity of up to one kilometer (Zonge and Hughes, 1988; Hasan et al., 2024). When combined with empirically based methodologies, CSAMT becomes an even more powerful tool for studying the incredibly diverse topographical features.”

REVISED PARAGRAPH:

“A number of prior groundwater investigations have made use of geophysical techniques (Bentley and Gharibi, 2004; Yadav and Singh, 2007; Fu et al., 2013; Vouillamoz et al., 2014; Robinson et al., 2016; Lin et al., 2018; Kouadio et al., 2020; Abbas et al., 2022; Kouadio et al., 2023; Zhang et al., 2024). A number of studies have shown that geophysical procedures outperform drilling techniques in terms of speed, ease of use, cost, and lack of invasiveness (Hu et al., 2013; Lin et al., 2018; Di et al., 2020; Fusheng et al., 2022; Hasan et al., 2024). Additionally, they are capable of conducting thorough geological evaluations in both the vertical and horizontal planes (Fu et al., 2013; Hasan et al., 2021). These methods are superior to others when it comes to collecting hydrogeological data from various subterranean habitats (Niwas and De Lima, 2003; Wynn et al., 2016; Kouadio et al., 2023). Groundwater studies nowadays often include resistivity surveys. Resistivity methods offer a broader resistivity range compared to other geophysical parameters, which is a major advantage (Bentley and Gharibi, 2004; Camporese et al., 2011; Robinson et al., 2016). The three main methods for measuring resistivity are the controlled source audio-frequency magnetotellurics (CSAMT), vertical electoral soundings (VES), and electrical resistivity tomography (ERT) (Soupios et al., 2007; Di et al., 2020; Zhang et al., 2024). Niwas and De Lima (2003), Soupios et al. (2007), Majumdar and Das (2011), Nwosu et al. (2013), Hasan et al. (2021), and Asfahani (2023) are among the previous groundwater-based geophysical studies that primarily utilized the VES method to evaluate groundwater resources in a single dimension. In recent decades, electrical resistivity tomography (ERT) has been widely used in hydrogeological studies for 2D and 3D assessment of groundwater resources (Bentley and Gharibi, 2004; Camporese et al., 2011; MLin et al., 2018; Abbas et al., 2022). However, it is unusual to evaluate aquifer yield at great depths in hard rock terrains using two and three dimensional hydraulic properties. Recent studies have demonstrated that CSAMT, which aims to gather extensive subsurface data at very deep depths using 2D/3D evaluations, is the most appropriate geophysical method for researching hard rock (Smith and Booker, 1991; Simpson and Bahr, 2005; Bai et al., 2010; Fu et al., 2013; Hu et al., 2013; Wang et al., 2015; Wynn et al., 2016; Zhang et al., 2021; Kouadio et al., 2023). The selection of resistivity methods in groundwater studies depends on several key factors, including survey objectives, depth of investigation, resolution, geological complexity, logistical constraints, cost and accessibility, electrical conductivity contrast, and field conditions, etc (Di et al., 2020; Hasan et al., 2024). VES is more suitable for shallow depths (< 200 m) for 1D resistivity imaging, has limited lateral resolution, is useful for layered aquifer characterization, works well in horizontally stratified formations, requires minimal equipment and is quick to deploy, is the most economical for small-scale studies, and may face limitations in highly resistive or conductive terrains (Soupios et al., 2007; Nwosu et al., 2013; Hasan et al., 2021). ERT offers better resolution for both shallow and intermediate depths (up to ~300 m) with 2D/3D imaging, provides high-resolution subsurface imaging, is ideal for detecting fractures/faults and heterogeneous aquifers, is better for complex geology (e.g., fractured zones, karst systems), is ideal for detailed aquifer geometry and contamination studies, needs more field effort and electrode spacing adjustments, and may face limitations in highly resistive or conductive terrains like VES (Camporese et al., 2011; MLin et al., 2018; Abbas et al., 2022). CSAMT is effective for deeper investigations (hundreds to thousands of meters) due to its low-frequency signal penetration, provides 2D/3D imaging over big area at large depths, has lower resolution than ERT but excels in deep structural mapping, is preferred for deep-seated structures like basement aquifers or geothermal systems, is used for regional groundwater exploration, demands specialized equipment and is more time-consuming compared with VES and ERT, is relatively more expensive than VES and ERT due to advanced instrumentation and processing, all resistivity methods rely on resistivity contrasts but CSAMT is more sensitive to deep conductive zones, and performs better in areas with cultural noise (e.g., urban settings) due to controlled signal sources (Zonge and Hughes, 1988; An et al., 2016; Kouadio et al., 2020; Zhang et al., 2021). When combined with empirically based methodologies, CSAMT becomes an even more powerful tool for studying the incredibly diverse topographical features at large depths (Hasan et al., 2024). So based on above factors, CSAMT was the most suitable method for this study.”

Comment 2:

Line 120- 138: the author might consider reorganizing this paragraph to make the significance of the resistivity method stand out.

Response 2:

The paragraph previously located in Lines 120–138 has been reorganized for improved clarity in the revised manuscript. For reference, both the original and the revised versions of the paragraph are provided below.

ORIGINAL PARAGRAPH:

“Several factors, such as the type of rock, fault, weathering degree, fluid content, permeability, pore-spacing, fracture, lithology, saturation, and joints, as well as the same structural heterogeneities, determine the geophysical and aquifer characteristics (Singh, 2005; Sinha et al., 2009; Hasan et al., 2021). Several prior studies utilized geophysical parameters in conjunction with hydraulic data or lithological logs to characterize underlying rock mass units hydrogeologically (De Lima and Niwas, 2000; Hubbard and Rubin, 2002; Niwas and De Lima, 2003; Singh, 2005; Soupios et al., 2007; Sinha et al., 2009; Majumdar and Das, 2011; Nwosu et al., 2013; Hasan et al., 2021; Asfahani, 2023). Resistivity methods provide an alternate option for aquifer parameter estimation by creating a beneficial relationship between electrical resistivity and the aquifer parameters (obtained from drilling tests). An innovative aspect of this work is its use of non-invasive geophysical techniques to create two- and three-dimensional k models in a diverse environment with a variety of rock types and significant depths. The planned study will necessitate the boring of a handful of boreholes at key spots all around the project site. A more trustworthy CSAMT study will allow us to evaluate the extensive research area. Then, by directly connecting geophysical and borehole data, k can be established for the entire researched site, even without drilling tests. Two- and three-dimensional k models are generated by applying the resulting equations to the full study area. This approach would reduce the need for costly boreholes to obtain a thorough and complete evaluation of subsurface hydrogeological conditions.”

REVISED PARAGRAPH:

“Resistivity methods are highly significant in groundwater studies due to their ability to characterize subsurface formations and identify potential aquifers. These techniques measure the electrical resistivity of subsurface materials, which varies depending on lithology, rock type, porosity, fluid content, weathering degree, faults, fractures, joints, saturation, and salinity (Singh, 2005; Sinha et al., 2009; Hasan et al., 2021). By integrating resistivity data with geological and hydrogeological information, researchers have been able to optimize the placement of wells, accurately assess groundwater potential, and support the effective and sustainable management of water resources (De Lima and Niwas, 2000; Hubbard and Rubin, 2002; Niwas and De Lima, 2003; Singh, 2005; Soupios et al., 2007; Sinha et al., 2009; Majumdar and Das, 2011; Nwosu et al., 2013; Hasan et al., 2021; Asfahani, 2023). By establishing a useful correlation between electrical resistivity and the limited borehole-based hydraulic parameters, resistivity methods offer the best alternative of the expensive drilling tests for estimating aquifer parameters over large area from shallow to large depths. In this study, we present a novel application of the Controlled Source Audio-Frequency Magnetotelluric (CSAMT) method to develop high-resolution 2D and 3D permeability (k) models at significant depths (up to 1300 meters) within a geologically complex and heterogeneous setting composed of sandstone, granite, and hornstone. Initially, a limited number of boreholes were drilled at strategically selected key locations across the study area. Subsequently, multiple CSAMT survey profiles were conducted, covering the entire region, including the borehole sites. By correlating the resistivity data obtained from the CSAMT surveys with permeability values derived from the borehole data, we established a reliable empirical relationship between resistivity and permeability. This correlation was then applied across the full CSAMT dataset, enabling the generation of 2D and 3D permeability models even in areas lacking direct borehole information. This approach allows for a more comprehensive and cost-effective assessment of deep groundwater resources, significantly reducing the need for extensive and expensive drilling campaigns.”

Comment 3:

Line 139- 140: the statement is too arbitrary; the language can be modified or more evidence is provided.

Response 3:

ORIGINAL STATEMENT:

“No one had ever tried to estimate K using direct or indirect methods in such a heterogeneous context before this work, where a broad diversity of rock types are present at a depth of 1 kilometer”.

REVISED STATEMENT:

“Prior to this study, no attempts had been made to estimate permeability (K) using either direct methods, such as borehole testing, or indirect geophysical approaches in such a geologically heterogeneous setting, characterized by a diverse mixture of sandstone, granite, and hornstone, extending to depths of up to one kilometer.”

Comment 4:

Line 161: in section 2.1 Study area, the author might consider adding more details about the rocks and geology of the study area.

Response 4:

Additional details about the rocks and geological characteristics of the study area have been added, as shown below.

ORIGINAL:

“Intruding rocks from the Indosinian, Caledonian, and Yanshanian eras are among the many geological formations and periods represented in the study region. Other layers from the Paleogene period are also present. The most common types of rock that have been discovered are sandstone, granite, and hornstone. The complex Kaiping concave fault and fold systems were the dominant geological features in the project region, which were developed as a result of magmatic processes and various structures (Qin, 2017). Emergence of joint fissured features symbolizes the various tectono-geological periods, with the local tectonic line corresponding with the faults strike, especially in the northeast orientation (Yang et al., 2021)”.

REVISED:

“The study area exhibits a complex and diverse geological history, characterized by well-defined geometrical relationships among various lithologies. These formations and structural features are the result of multiple tectono-magmatic events spanning several geological periods. Intrusive rocks from the Indosinian (Late Triassic), Caledonian (Silurian–Devonian), and Yanshanian (Jurassic–Cretaceous) orogenies are well-represented, indicating a long sequence of crustal deformation and magmatic activity. These intrusions are primarily composed of granitic bodies, which suggest deep-seated magmatic processes associated with continental collision and subduction zones. In addition to these intrusive phases, sedimentary strata from the Paleogene period are also present, reflecting a later stage of basin development with fluvial and lacustrine depositional environments. Among the most prevalent rock types encountered in the region are sandstone, granite, and hornstone. Sandstone reflects high-energy sedimentary deposition. Granite indicates deep magmatic intrusions likely associated with Yanshanian tectonics. Hornstone (hornfels) results from contact metamorphism caused by magma intruding sedimentary rocks. The structural framework of the region is dominated by the Kaiping concave fault and fold system, a geologically significant and highly deformed zone that reflects multiple deformation episodes (Qin, 2017). These structures were primarily shaped by magmatic intrusions, crustal movements, and regional stress regimes. The presence of extensive jointed and fissured zones throughout the rock mass further supports a history of dynamic tectonic activity. These joints often serve as secondary permeability pathways and are critical in controlling groundwater flow in the fractured rock environment. Importantly, the orientation of these structural features, including faults and joints, is often aligned with northeast-trending tectonic lines, which are consistent with broader regional stress directions (Yang et al., 2021). This relationship among lithologies and structural features plays a critical role in controlling groundwater flow and permeability distribution.”

Comment 5:

Line 204: what does “5-20%” represent for? More specific content is preferred for this sentence.

Response 5:

“The vertical resolution of 5–20% can be assessed by CSAMT when exploring depths ranging from 20 to 1000 meters” explained as in the revised version:

“In CSAMT, the vertical resolution, which refers to the ability to distinguish between adjacent subsurface layers, can typically range between 5% and 20% of the investigation depth from approximately 20 to 1000 meters. At shallower depths (e.g., 20–100 m), vertical resolution is higher (closer to 5%), enabling better differentiation between thin layers. At greater depths (up to 1000 m), resolution may degrade toward the 20% mark due to signal attenuation and broader averaging of resistivity data. This makes CSAMT a valuable tool for identifying significant lithological contrasts, fault zones, and resistivity anomalies related to geological structures”.

Comment 6:

Line 217: in section 2.2.2, why were 6 profiles selected? How did the author determine the locations of the profiles? The author might consider providing more evidence of the site location and data collection in the supporting material.

Response 6:

Further details regarding the selection criteria and rationale for the survey profiles have been provided in the revised manuscript.

ORIGINAL:

“The CSAMT data was acquired using six profiles (1–6) with a 50 meter interval between each station”.

REVISED:

“CSAMT data were acquired along six profiles (profiles 1–6), with a station spacing of 50 meters between each measurement point. The location of 6 CSAMT profiles was chosen based on several factors, including geological targets and objectives, surface geology and mapping data, topography and terrain accessibility, orientation relative to structures, spacing and coverage requirements, resistivity contrast expectations, integration with other data (boreholes), environmental and regulatory constraints, and source-receiver geometry requirements, etc. Carefully selected survey profiles enhanced the ability to resolve critical subsurface features and minimized ambiguities in the geophysical interpretation”.

Additional details are provided in the revised manuscript.

Comment 7:

Line 250-154: The author might consider providing more details of the static correction and the Hanning window spatial filtering method.

Response 7:

Additional details on static correction and Hanning window spatial filtering have been included in the revised manuscript to enhance clarity and support the interpretation of CSAMT data.

ORIGINAL:

“The static corrections were made using a Hanning window spatial filtering method, which involved geological information and curve analysis.”

REVISED:

“Static correction and spatial filtering using a Hanning window are essential preprocessing steps in CSAMT data analysis, aimed at improving data quality and enhancing the reliability of subsurface resistivity models. Static correction addresses the effects of near-surface resistivity inhomogeneities, which can distort electric field measurements and introduce static shifts, vertical displacements in apparent resistivity curves that misrepresent deeper subsurface conditions. This correction typically involves adjusting the measured electric fields by referencing them to a stable or averaged field, effectively removing shallow-layer influences and isolating true subsurface signals. Spatial filtering, on the other hand, is used to mitigate noise introduced by environmental and instrumental sources. Among various filters, the Hanning (Hann) window is commonly applied due to its effectiveness in reducing spectral leakage and smoothing data. When used in spatial filtering, the Hanning window averages measurements across adjacent stations in a weighted manner, preserving spatial trends while suppressing high-frequency noise. This improves the coherence of the dataset and ensures more stable and interpretable inversion results”.

Comment 8:

Figure 1: typos in (b), “uncertainty”; also the words are too small to read.

Response 8:

Figure 1, along with all other figures, has been redrawn and improved for better clarity and presentation. The updated figs are included at the end of the response/comments section in the attached file.

Comment 9:

Figure 7 and Figure 8: a little confused about the legend of the north direction in both figures

Response 9:

The north direction in these figures is correctly oriented, though slightly tilted, to provide a clearer and more informative view of the 3D permeability (k) models. The revised figures are included at the end of the response/comments section in the attached file.

Referee Comments (RC1):

Hasan and Su tried to get the permeability of deep hard rock by a geophysical approach. They used the CSAMT to get the resistivity of the subsurface for several 2D profiles and then built an empirical equation to describe the relationship between permeability and resistivity using 37 permeabilities from boreholes. Then they applied the equation to all 2D profiles and finally extrapolated 2D profiles to a 3D field. They argued that the advantage of their approach is that it is able to get the information of the deep subsurface to 1 km depth. They also said their approach is low-cost than borehole drilling. I am sorry I am not that encouraging to this work. I don’t think this work is novel enough for publication in HESS, a top journal in hydrology and water resources. I suggest rejecting it for publication in other journals of geology. Following are my reasons:

Response:

Dear Anonymous Reviewer,

Thank you for your detailed and thoughtful review of our manuscript. We sincerely appreciate the time and effort you dedicated to evaluating our work.

We greatly appreciate HESS's reputation as a leading journal in the fields of hydrology and water resources. As groundwater assessment inherently depends on a comprehensive understanding of subsurface geology, we believe that the integration of geophysical methods, particularly in complex geological settings, adds significant value to the scope of the journal.

This work builds upon our extensive experience in applying geophysical techniques to groundwater hydrology, with previous publications in the high-impact journals. Given the focus and scientific standards of HESS, as well as the novelty of our current approach, we believe that this manuscript aligns well with the journal’s objectives.

It is worth noting that the handling editors initially assessed the novelty and relevance of our submission before inviting peer review. Additionally, the other reviewers have recommended only minor revisions and expressed strong support for the publication of this work, further reinforcing its contribution and suitability for HESS.

We acknowledge that many of your comments are highly constructive and have been very helpful in improving our manuscript. These include your observation regarding the correct unit of permeability (mD, not m/d), your suggestion to expand the dataset, and your recommendation to discuss theoretical relationships such as the Archie and Kozeny-Carman equations, etc. All of these aspects have been carefully addressed in the revised version.

However, we also feel that certain criticisms may have overlooked the broader context and significance of our contributions, especially when compared with the existing body of literature. Although many studies in the past have explored the estimation of permeability (k) or other hydraulic parameters using geophysical data, this work introduces several key innovations that distinguish it from previous research.

When compared to existing literature, the following contributions represent significant advancements:

1. First-time measurement of permeability (k) at depths exceeding 1 km in a hard-rock environment.
2. Generation of 2D and 3D k models, to our knowledge, the first of their kind for deep groundwater assessments using any geophysical method.
3. Innovative use of CSAMT to derive volumetric estimations of hydraulic parameters such as permeability, which has not been attempted previously.
4. Application in a highly heterogeneous lithological context involving sandstone, granite, and hornstone, where such estimation of k had never been conducted before.
5. Efficient integration of sparse borehole data to produce high-resolution subsurface models over a large hard-rock terrain, demonstrating a practical alternative to intensive drilling.
6. Significant reduction in the need for costly deep boreholes, which would otherwise be required in the hundreds to achieve a similar level of spatial resolution.

We hope these clarifications help illustrate the novelty and importance of this work, and we remain grateful for your review, which ultimately contributed to strengthening our manuscript. We look forward to the opportunity to share this contribution with the HESS readership.

 Comment 1:

Either component of this work is very old. The CSAMT can date back to 1970s. The relationship between the permeability and resistivity has also been largely studied so far. The inversion of permeability/hydraulic conductivity using geophysical approaches is so well-known.

Response 1:

We thank the reviewer for this important comment.

We agree that both CSAMT and the general concept of relating resistivity to permeability have a long history in geophysical and hydrogeological research. However, the key contribution of our study lies not in the introduction of entirely new methods, but in the novel integration and application of established techniques to a unique and challenging geological context, resulting in unprecedented outcomes.

Specifically, while CSAMT has traditionally been used for deep geophysical investigations, its application in hydrogeology for constructing 2D and 3D models of permeability (k) at depths exceeding 1 km, especially in a highly heterogeneous hard-rock setting comprising sandstone, granite, and hornstone, has not been demonstrated in any previous work, to the best of our knowledge.

In fact, no past studies have reported the generation of volumetric (2D and 3D) models of k or any hydraulic parameter using either direct (drilling-based) or indirect (geophysical, etc) methods at such depths and across such varied lithologies. In our earlier work, we demonstrated for the first time the use of ERT to generate shallow-depth 2D and 3D models of hydraulic conductivity in a tuff rock environment. Building on that, this study extends the concept both methodologically and spatially by using CSAMT to model permeability at much greater depths and over geologically complex terrain.

Furthermore, we have now included a more detailed discussion in the revised manuscript regarding the relationship between resistivity and permeability (e.g., using the Archie and Kozeny-Carman equations), along with how our approach advances existing methodologies.

We hope this clarification better highlights the novelty, scale, and practical significance of our work within the context of existing literature.

Comment 2:

Authors even didn’t well discuss the previous studies of relationship between permeability and resistivity, such as the Archie equation and Kozeny-Carman equation.

Response 2:

We thank the reviewer for this insightful comment. In response, we have included a detailed discussion in the revised manuscript regarding previous studies on the relationship between resistivity and permeability, specifically addressing the Archie equation and the Kozeny-Carman equation. These models are fundamental in petrophysical and hydrogeophysical studies and serve as a theoretical basis for linking electrical resistivity measurements to hydraulic properties. Their relevance, limitations in heterogeneous geological settings, and how our approach builds upon these foundations are now clearly articulated in the revised version of the manuscript.

A brief discussion is given below:

Discussion on the Relationship between Permeability and Resistivity

Several foundational studies have established empirical and theoretical relationships between electrical resistivity and hydraulic properties such as permeability. The Archie equation, introduced by Archie (1942), is widely used in clean, saturated sedimentary formations. It relates formation resistivity to porosity and water saturation but assumes the absence of clay minerals and thus has limitations in more complex lithologies.

The Kozeny-Carman equation is another widely accepted model that links permeability to porosity and specific surface area. While it does not directly involve resistivity, it is often used alongside petrophysical models to interpret hydrogeological characteristics based on geophysical data.

Although these relationships are well-established, their application is often restricted to homogeneous or semi-homogeneous geological settings. In contrast, our study extends the use of these principles into a highly heterogeneous hard rock context, including sandstone, granite, and hornstone, at depths exceeding 1 km. By integrating borehole-derived permeability with CSAMT-based resistivity data, we derive a site-specific empirical model that enables the construction of 2D and 3D permeability distributions across the entire study area, an approach not previously demonstrated in the literature.

Archie’s Equation

Archie’s law (Archie, 1942) relates the bulk electrical resistivity of a fully saturated porous medium to its porosity and fluid resistivity. It is commonly expressed as:

where:

ρb​ is the bulk resistivity,

ρf​ is the fluid resistivity,

ϕ is the porosity,

a and m are empirical constants.

Although Archie’s law does not directly estimate permeability, porosity is often used as a proxy because of its influence on fluid flow. The resistivity-porosity relationship can be indirectly extended to infer permeability, especially when combined with other petrophysical models.

Kozeny-Carman Equation

The Kozeny-Carman equation provides a direct relationship between permeability (kkk) and porosity and is given by:

where:

k is the permeability,

ϕ is the porosity,

S is the specific surface area,

C is a constant related to pore structure and tortuosity.

By combining the Kozeny-Carman equation with Archie’s law, researchers have developed empirical and semi-empirical models to relate geophysical measurements (like resistivity) to permeability.

Relevance to This Study

While these classical models provide important theoretical underpinnings, their direct application in complex geological environments, such as heterogeneous hard rock formations (e.g., granite, sandstone, hornstone), is often limited due to variability in mineralogy, pore structure, and anisotropy. In this study, we established an empirical relationship between resistivity and permeability based on field measurements from boreholes and CSAMT profiles. This site-specific correlation enables the generation of 2D and 3D permeability models in geologically complex settings where traditional models may fall short.

Comment 3:

A perfect equation (Fig. 3) was built using only 37 known permeabilities. It is hard to convince me the number of k values is enough to build an equation for a domain of 1.8 km*1.8 km*1km of high heterogeneity. It is hard to say this equation is still available for other positions and depths in the domain. As the authors also mentioned, the resistivity is determined by many other factors, not only the permeability.

Response 3:

We thank the reviewer for this important comment regarding the validity and representativeness of the empirical relationship established in Figure 3. We fully agree that both resistivity and permeability (k) are influenced by multiple geological and physical factors, such as lithology, degree of weathering, fluid content, porosity, fracturing, saturation, and jointing, among others. This complexity underscores the need for careful calibration of any empirical relationship.

As mentioned in the original manuscript, the initial empirical equation (R² = 0.97) was built using 37 carefully selected borehole-derived k values. These were not arbitrarily chosen; rather, they were distributed across all three major rock types in the area, sandstone, granite, and hornstone, and covered the full observed range of resistivity (72–4765 Ω·m) and permeability (0.01–19.8 mD). Our intention was to ensure the equation’s applicability across the domain by capturing the full variability of both geological conditions and resistivity-permeability values within the study area.

To further address the concern about sample size and enhance the robustness of our model, we have now expanded the dataset by adding 79 new data points, bringing the total to 116. This expanded dataset includes wider spatial coverage and continues to represent all three lithologies. The updated empirical equation based on 116 data points has a slightly adjusted but still strong correlation (R² = 0.96), confirming the reliability of the established relationship. The new data spans a resistivity range of 35–4765 Ω·m and a permeability range of 0.01–19.9 mD, reflecting a comprehensive coverage of the variability in the study domain (1.8 × 1.8 × 1.0 km).

The drilling locations were strategically selected to capture both spatial and geological heterogeneities, ensuring that the derived empirical model can be confidently applied to the entire domain—even at locations and depths where no direct permeability measurements were available. This approach significantly enhances the feasibility of deep permeability modeling in data-scarce regions and reduces the need for excessive drilling, which is costly and often impractical in hard rock terrains.

A more detailed explanation, including updated figures and statistical analyses, is provided in the revised manuscript. Please also refer to the attached file for the updated figures, particularly Figures 3 and 4, which now include additional data points to support the revised empirical model.

Comment 4:

The fitted line is too perfect to believe. I always saw the fitting as follows (very noisy):

Response 4:

We appreciate the reviewer’s observation and understand the concern regarding the quality of the curve fitting. However, the smoothness or “perfection” of a fitted relationship between resistivity and permeability (k) is highly dependent on a number of factors that vary significantly across different studies. These factors include the geological setting, the distribution and range of data points, the lithology, the degree of heterogeneity, and the accuracy of both resistivity and k measurements.

In our study, the resistivity values span a wide range (approximately 35 to 4765 Ω·m), and the permeability values range from 0.01 to 19.9 mD. This wide dynamic range helps to better resolve trends, especially in high-resistivity rocks such as granite, where permeability remains very low and changes minimally. In such cases, large differences in resistivity correspond to small variations in permeability, naturally resulting in a smoother inverse curve.

We agree that in many past studies, the relationship appears noisy, often because the resistivity data is concentrated in a narrower range (e.g., 50–300 Ω·m as shown in the figs provided by the reviewer), and the variations in k may be more influenced by local heterogeneities. In contrast, the broader data spread in our study, combined with careful selection of measurement locations to capture the variability of all three dominant lithologies (sandstone, granite, and hornstone), contributes to a more stable empirical trend.

In the revised manuscript, we have increased the number of data points from 37 to 116, yet the curve still maintains a high degree of correlation (R² = 0.96), suggesting the robustness of the relationship rather than overfitting. The updated Figures 3 and 4 (attached) show the expanded dataset and reaffirm this trend.

Additionally, we have emphasized in the manuscript that both resistivity and k are influenced by a range of factors, such as porosity, saturation, fluid content, fractures, and lithology, which are inherently linked in this geological context. While some scatter exists, particularly at lower resistivity values (<1500 Ω·m), the overall trend aligns well with the theoretical and empirical basis established in previous research showing an inverse relationship between resistivity and permeability.

We hope this clarifies the rationale behind the nature of the fitted curve and addresses the reviewer’s concern.

A more detailed discussion on this point has been included in the revised manuscript to clarify the methodology and support the reliability of the derived empirical relationship.

Comment 5:

I am not sure the resolution of the resistivity obtained by CSAMT and the permeability obtained by pumping test. Do the scales match well? i.e., what’s the size of the pixel in the maps in Fig. 4? The k obtained by pumping test always represents the average hydraulic conductivity of an area, so do this range and your pixel size match?

Response 5:

Thank you for raising this important point regarding the consistency between the resolution of resistivity measurements from CSAMT and the permeability (k) values derived from borehole tests.

The accuracy of the estimated k values from our empirical relationship is primarily dependent on the quality of the input data: (1) the resistivity values derived from CSAMT and (2) the k values obtained from boreholes. In our study, we ensured a high-quality CSAMT dataset through optimized survey design, careful data acquisition, and advanced processing and inversion techniques. The resulting models achieved a root mean square (RMS) misfit of less than 5%, ensuring reliable subsurface resistivity measurements. Additional details on data acquisition and processing are provided in the revised manuscript.

Regarding k measurements, we acknowledge that pumping tests yield average hydraulic conductivity over relatively large volumes, which are suitable for 1D correlations with averaged geophysical data. However, our objective in this study was to construct high-resolution 2D and 3D permeability models, which require point-based permeability measurements that are more compatible with the spatial resolution of CSAMT-derived resistivity data. To achieve this, we relied on rock core tests rather than pumping tests. Rock core analysis allows for permeability measurements at specific depths and locations, enabling a more precise match with the local resistivity values at those points.

As for the model resolution, each pixel in the 2D maps represents an area of approximately 50 m × 50 m in the horizontal direction, which directly corresponds to the CSAMT station spacing. The vertical resolution varies with depth but generally falls within tens of meters across the total investigated depth of approximately 1300 m, depending on signal penetration and inversion sensitivity. This level of resolution is well-suited for resolving key subsurface features and provides a robust basis for the reliability of the derived permeability models.

Therefore, both the k values and resistivity measurements were acquired and processed at compatible spatial scales, ensuring that the empirical relationship and resulting 2D/3D k models are internally consistent and scientifically robust. Further explanation of this methodology is included in the revised version of the manuscript.

Comment 6:

Table 2, the same problem, even for the pumping test itself, the k values obtained are with large uncertainties, the deviations of one order of magnitude is not surprising. However, the differences of k values between the CSAMT and borehole drilling are less than 1 or 0.1, which is unbelievable.

Response 6:

Thank you for this valuable observation. We would like to clarify that Table 2 presents a comparison between the predicted and measured k values for only 12 selected data points. These points were chosen to highlight instances of close agreement; however, they do not represent the full variability in the dataset.

The reviewer noted that “the differences of k values between the CSAMT and borehole drilling are less than 1 or 0.1,” but this is not consistently true, even within the selected 12 data points. For instance, a difference of 3.6 mD and 2.7 mD was observed between predicted and measured k values for sounding number P6-1 and well 4 at depths of 10 m and 45 m, respectively.

Furthermore, many data points in the full dataset of 116 borehole locations exhibit larger differences between predicted and measured k values, and we have included a selection of these in the revised version of the manuscript to provide a more comprehensive view. These updates better reflect the natural uncertainty and variability in both field measurements and predictions.

Comment 7:

I am not sure why very deep k is necessary. A latest study found that groundwater deep than 500 m is not an active component in terrestrial hydrologic cycle and the water there might be brine. You may say your work found there are a lot of sandy rocks in depth. However, given the large uncertainties of your approach, did you compare the findings with the results of local or national geologic survey? DOI:10.1038/s43247-023-00697-6

Response 7:

Thank you for this insightful comment and for highlighting the referenced study.

While it is true that deep groundwater (below 500 m) is often less connected to the active terrestrial hydrologic cycle and may contain brine, the necessity for deep groundwater exploration in our study area arises from several critical and site-specific factors:

1. Surface water resources in the study area are limited and unreliable, making deep groundwater a potentially vital alternative water source.
2. Shallow zones are predominantly composed of fresh granite, which typically has low permeability and limited groundwater potential. In contrast, fractured granite, sandstone, and hornstone formations that can host significant groundwater resources are found at greater depths.
3. In China, recent national groundwater resource assessment initiatives have prioritized deep earth exploration in areas with potential for deep aquifers, to support sustainable development and ensure long-term water security, especially in regions experiencing severe water stress.
4. Deep groundwater investigations are essential for:

• Identifying hidden but critical water sources.
• Characterizing deep aquifers and understanding their storage and recharge potential.
• Supporting strategic water resource planning in response to increasing demand and climate variability.

To address concerns regarding uncertainty:

• We employed a robust CSAMT survey design, followed by accurate resistivity inversion and low-RMS models.
• Rock core testing was used to measure k values at multiple depths with high confidence.
• An empirical equation based on a representative and extensive dataset (116 points) was used to derive 2D and 3D k models.

Furthermore, we compared our findings with existing geological information from both local and national geological surveys, and found our results to be consistent with the known stratigraphy and hydrogeological features of the region. Additional information on this comparison has been included in the revised manuscript.

We believe that, despite the challenges, this deep groundwater investigation provides valuable insights and practical relevance, especially in arid and semi-arid regions where shallow resources are scarce or overexploited.

A more detailed explanation and supporting discussion on this point has been incorporated into the revised version of the manuscript.

Comment 8:

I found a work of the authors just published in scientific reports https://www.nature.com/articles/s41598-025-85626-7 It is exactly the same workflow with this one. Many sentences are the same. I am not supportive for research of such a style in the community.

Response 8:

We appreciate the reviewer’s observation and the opportunity to clarify the distinction between the two works.

The paper recently published in Scientific Reports (DOI: 10.1038/s41598-025-85626-7) focuses on the evaluation of site suitability for the installation of China’s Next Generation Neutrino Detector, the Jiangmen Underground Neutrino Observatory (JUNO). That study utilizes CSAMT-derived geomechanical properties, specifically the Rock Quality Designation (RQD), to assess the rock mass stability and integrity for large-scale underground construction.

In contrast, the current manuscript focuses on a hydrogeological application, using CSAMT and borehole core data to derive permeability (k) models for deep groundwater resource assessment. While both studies are conducted within the broader Kaiping region of South Guangdong, a region of high scientific and strategic interest characterized by complex geological heterogeneity, their objectives, methodologies, and scientific contributions are fundamentally different.

We acknowledge that there may be some overlap in structure due to the use of similar geophysical techniques in similar geological settings. However, the purpose and interpretation of CSAMT data in each study differ significantly:

• In the Scientific Reports paper, CSAMT data were used to evaluate mechanical stability (RQD) for a construction project.
• In the present study, CSAMT data are employed to estimate hydraulic conductivity (k) and to develop 2D and 3D permeability models relevant to deep groundwater systems.

In response to this concern and to ensure the novelty of our current submission, we have:

1. Redrawn all figures to reflect the unique objectives of this study.
2. Substantially revised the entire manuscript text, including the introduction, methodology, results/discussion, and conclusions to emphasize the hydrogeological focus.
3. Incorporated reviewer suggestions and community feedback to further distinguish this work from previous publications.

We would also like to highlight that both papers are part of different national-level projects, each addressing unique scientific challenges: one in geotechnical engineering, and the other in deep groundwater exploration. These investigations are aligned with China's broader strategy for deep subsurface resource development.

We respectfully believe that the current manuscript stands as an original and independent contribution, both in scope and significance, and adds value to the scientific understanding of subsurface hydrogeological processes.

Comment 9:

The structure of the manuscript is OK, however, the writing is still unshaped. Many sentences are duplicated, such as lines 306-308 which appear many times. I think once you clarified the strengths of your approaches in the introduction, you only need to describe you approach in methodology and it is redundant to say this again. Also line 218 “About 1300 meters was the depth of investigation (DOI) in the CSAMT investigation.”, such an expression is awkward. Why not “the depth of investigation (DOI) in the CSAMT investigation was about 1300 meters”.

Response 9:

We thank the reviewer for their constructive feedback regarding the manuscript’s writing quality and structure.

In response, we have carefully revised and reshaped the manuscript in accordance with the suggestions provided. Specifically:

• All duplicate or repetitive sentences, including those previously appearing in multiple sections (e.g., lines 306–308), have been removed or appropriately rephrased to avoid redundancy.
• We have streamlined the narrative by presenting the strengths of our approach clearly in the Introduction and limiting their repetition in subsequent sections.
• As suggested, we have improved awkward or unclear expressions, such as line 218. The sentence now reads:
  
  “The depth of investigation (DOI) in the CSAMT survey was approximately 1300 meters.”

We believe these revisions enhance the overall clarity, coherence, and professionalism of the manuscript and we appreciate the reviewer’s guidance in helping us improve the quality of the submission.

Comment 10:

What are you trying to get? Permeability or hydraulic conductivity? I don’t think the unit of permeability is L/T.

Response 10:

We appreciate the reviewer’s comment and the opportunity to clarify this point.

Our study focuses on permeability, and the correct unit is milliDarcy (mD). In the initial version of the manuscript, permeability was mistakenly written using the unit m/d, which may have caused confusion with hydraulic conductivity, typically expressed in length per time (L/T) units such as m/day.

To address this:

• The methodology section has been revised to clearly state that permeability is measured in mD.
• This correction has been applied consistently throughout the manuscript and all figures.

We thank the reviewer for catching this important detail, which has now been rectified in the revised version.

Referee Comments (RC2):

This paper addresses the challenges associated with current groundwater exploration and evaluates the advantages and disadvantages of various methods for measuring hydraulic parameters. The author highlights the application of a novel approach, by using Controlled Source Audio-Frequency Magnetotellurics (CSAMT) method, which is employed to estimate 2D and 3D permeability at depths exceeding 1 km in highly heterogeneous rock environments. The study presents its methods and findings in a well-structured manner, offering insights into deep groundwater exploration.

However, certain assertions appear overly generalized and could benefit from further substantiation. Additionally, more detailed descriptions of the methodologies and the study area would enhance the clarity, reproducibility, and robustness of the research.

Response:

We sincerely thank the anonymous reviewer for their insightful and constructive feedback, which has significantly contributed to enhancing the quality of our work. We have made every effort to revise the manuscript thoroughly in line with the reviewer’s suggestions.

In the revised version, we have expanded the descriptions of both the study area and the methodologies to provide greater clarity and context. Additional explanations and supporting information have also been included to substantiate our findings and address the points raised.

As per the journal’s submission guidelines, we are first submitting our detailed responses to the reviewer’s comments. Following this, we will submit the revised manuscript reflecting all the suggested changes.

Specific comments:

Comment 1:

Line108-119: the author might consider adding a little more evidence of the reason that CSAMT is selected for this study. For example, the author stated that VES method is used to evaluate groundwater resources in a single dimension by a broad of previous studies, but did not illustrate the background about why they did not use other methods, like CSAMT or ERT. Additionally, the author states that there are three main methods, but there are very few examples or introductions about ERT in this paragraph.

Response 1:

The entire paragraph from Lines 88–119 has been revised for improved clarity and structure in the updated manuscript. Both the original and the revised versions of the paragraph are provided below for comparison.

ORIGINAL PARAGRAPH:

“A number of prior groundwater investigations have made use of geophysical techniques (Bentley and Gharibi, 2004; Yadav and Singh, 2007; Fu et al., 2013; Vouillamoz et al., 2014; Robinson et al., 2016; Lin et al., 2018; Kouadio et al., 2020; Abbas et al., 2022; Kouadio et al., 2023; Zhang et al., 2024). A number of studies have shown that geophysical procedures outperform drilling techniques in terms of speed, ease of use, cost, and lack of invasiveness (Hu et al., 2013; Lin et al., 2018; Di et al., 2020; Fusheng et al., 2022; Hasan et al., 2024). Additionally, they are capable of conducting thorough geological evaluations in both the vertical and horizontal planes (Fu et al., 2013; Hasan et al., 2021). These methods are superior to others when it comes to collecting hydrogeological data from various subterranean habitats (Niwas and De Lima, 2003; Wynn et al., 2016; Kouadio et al., 2023). Groundwater studies nowadays often include resistivity surveys. Resistivity methods offer a broader resistivity range compared to other geophysical parameters, which is a major advantage (Bentley and Gharibi, 2004; Camporese et al., 2011; Robinson et al., 2016). The three main methods for measuring resistivity are the controlled source audio-frequency magnetotellurics (CSAMT), vertical electoral soundings (VES), and electrical resistivity tomography (ERT) (Soupios et al., 2007; Di et al., 2020; Zhang et al., 2024). Niwas and De Lima (2003), Soupios et al. (2007), Majumdar and Das (2011), Nwosu et al. (2013), Hasan et al. (2021), and Asfahani (2023) are among the previous groundwater-based geophysical studies that primarily utilized the VES method to evaluate groundwater resources in a single dimension. It is unusual to evaluate aquifer yield at great depths in hard rock terrains using two- and three-dimensional hydraulic properties. Recent studies have demonstrated that CSAMT, which aims to gather extensive subsurface data at very deep depths using 2D/3D evaluations, is the most cost-effective and appropriate geophysical method for researching hard rock (Smith and Booker, 1991; Simpson and Bahr, 2005; Bai et al., 2010; Fu et al., 2013; Hu et al., 2013; Wang et al., 2015; Wynn et al., 2016; Di et al., 2020; Zhang et al., 2021; Kouadio et al., 2023; Hasan et al., 2024). Advantages of CSAMT over other geophysical research methods include its lower cost, its responsiveness to low-resistance rocks, and its ease of usage in challenging topographic circumstances (An et al., 2016; Kouadio et al., 5 115 2020; Zhang et al., 2021). Compared to most geophysical technologies, including ERT, CSAMT's subsurface assessment capabilities are superior due to its depth capacity of up to one kilometer (Zonge and Hughes, 1988; Hasan et al., 2024). When combined with empirically based methodologies, CSAMT becomes an even more powerful tool for studying the incredibly diverse topographical features.”

REVISED PARAGRAPH:

“A number of prior groundwater investigations have made use of geophysical techniques (Bentley and Gharibi, 2004; Yadav and Singh, 2007; Fu et al., 2013; Vouillamoz et al., 2014; Robinson et al., 2016; Lin et al., 2018; Kouadio et al., 2020; Abbas et al., 2022; Kouadio et al., 2023; Zhang et al., 2024). A number of studies have shown that geophysical procedures outperform drilling techniques in terms of speed, ease of use, cost, and lack of invasiveness (Hu et al., 2013; Lin et al., 2018; Di et al., 2020; Fusheng et al., 2022; Hasan et al., 2024). Additionally, they are capable of conducting thorough geological evaluations in both the vertical and horizontal planes (Fu et al., 2013; Hasan et al., 2021). These methods are superior to others when it comes to collecting hydrogeological data from various subterranean habitats (Niwas and De Lima, 2003; Wynn et al., 2016; Kouadio et al., 2023). Groundwater studies nowadays often include resistivity surveys. Resistivity methods offer a broader resistivity range compared to other geophysical parameters, which is a major advantage (Bentley and Gharibi, 2004; Camporese et al., 2011; Robinson et al., 2016). The three main methods for measuring resistivity are the controlled source audio-frequency magnetotellurics (CSAMT), vertical electoral soundings (VES), and electrical resistivity tomography (ERT) (Soupios et al., 2007; Di et al., 2020; Zhang et al., 2024). Niwas and De Lima (2003), Soupios et al. (2007), Majumdar and Das (2011), Nwosu et al. (2013), Hasan et al. (2021), and Asfahani (2023) are among the previous groundwater-based geophysical studies that primarily utilized the VES method to evaluate groundwater resources in a single dimension. In recent decades, electrical resistivity tomography (ERT) has been widely used in hydrogeological studies for 2D and 3D assessment of groundwater resources (Bentley and Gharibi, 2004; Camporese et al., 2011; MLin et al., 2018; Abbas et al., 2022). However, it is unusual to evaluate aquifer yield at great depths in hard rock terrains using two and three dimensional hydraulic properties. Recent studies have demonstrated that CSAMT, which aims to gather extensive subsurface data at very deep depths using 2D/3D evaluations, is the most appropriate geophysical method for researching hard rock (Smith and Booker, 1991; Simpson and Bahr, 2005; Bai et al., 2010; Fu et al., 2013; Hu et al., 2013; Wang et al., 2015; Wynn et al., 2016; Zhang et al., 2021; Kouadio et al., 2023). The selection of resistivity methods in groundwater studies depends on several key factors, including survey objectives, depth of investigation, resolution, geological complexity, logistical constraints, cost and accessibility, electrical conductivity contrast, and field conditions, etc (Di et al., 2020; Hasan et al., 2024). VES is more suitable for shallow depths (< 200 m) for 1D resistivity imaging, has limited lateral resolution, is useful for layered aquifer characterization, works well in horizontally stratified formations, requires minimal equipment and is quick to deploy, is the most economical for small-scale studies, and may face limitations in highly resistive or conductive terrains (Soupios et al., 2007; Nwosu et al., 2013; Hasan et al., 2021). ERT offers better resolution for both shallow and intermediate depths (up to ~300 m) with 2D/3D imaging, provides high-resolution subsurface imaging, is ideal for detecting fractures/faults and heterogeneous aquifers, is better for complex geology (e.g., fractured zones, karst systems), is ideal for detailed aquifer geometry and contamination studies, needs more field effort and electrode spacing adjustments, and may face limitations in highly resistive or conductive terrains like VES (Camporese et al., 2011; MLin et al., 2018; Abbas et al., 2022). CSAMT is effective for deeper investigations (hundreds to thousands of meters) due to its low-frequency signal penetration, provides 2D/3D imaging over big area at large depths, has lower resolution than ERT but excels in deep structural mapping, is preferred for deep-seated structures like basement aquifers or geothermal systems, is used for regional groundwater exploration, demands specialized equipment and is more time-consuming compared with VES and ERT, is relatively more expensive than VES and ERT due to advanced instrumentation and processing, all resistivity methods rely on resistivity contrasts but CSAMT is more sensitive to deep conductive zones, and performs better in areas with cultural noise (e.g., urban settings) due to controlled signal sources (Zonge and Hughes, 1988; An et al., 2016; Kouadio et al., 2020; Zhang et al., 2021). When combined with empirically based methodologies, CSAMT becomes an even more powerful tool for studying the incredibly diverse topographical features at large depths (Hasan et al., 2024). So based on above factors, CSAMT was the most suitable method for this study.”

Comment 2:

Line 120- 138: the author might consider reorganizing this paragraph to make the significance of the resistivity method stand out.

Response 2:

The paragraph previously located in Lines 120–138 has been reorganized for improved clarity in the revised manuscript. For reference, both the original and the revised versions of the paragraph are provided below.

ORIGINAL PARAGRAPH:

“Several factors, such as the type of rock, fault, weathering degree, fluid content, permeability, pore-spacing, fracture, lithology, saturation, and joints, as well as the same structural heterogeneities, determine the geophysical and aquifer characteristics (Singh, 2005; Sinha et al., 2009; Hasan et al., 2021). Several prior studies utilized geophysical parameters in conjunction with hydraulic data or lithological logs to characterize underlying rock mass units hydrogeologically (De Lima and Niwas, 2000; Hubbard and Rubin, 2002; Niwas and De Lima, 2003; Singh, 2005; Soupios et al., 2007; Sinha et al., 2009; Majumdar and Das, 2011; Nwosu et al., 2013; Hasan et al., 2021; Asfahani, 2023). Resistivity methods provide an alternate option for aquifer parameter estimation by creating a beneficial relationship between electrical resistivity and the aquifer parameters (obtained from drilling tests). An innovative aspect of this work is its use of non-invasive geophysical techniques to create two- and three-dimensional k models in a diverse environment with a variety of rock types and significant depths. The planned study will necessitate the boring of a handful of boreholes at key spots all around the project site. A more trustworthy CSAMT study will allow us to evaluate the extensive research area. Then, by directly connecting geophysical and borehole data, k can be established for the entire researched site, even without drilling tests. Two- and three-dimensional k models are generated by applying the resulting equations to the full study area. This approach would reduce the need for costly boreholes to obtain a thorough and complete evaluation of subsurface hydrogeological conditions.”

REVISED PARAGRAPH:

“Resistivity methods are highly significant in groundwater studies due to their ability to characterize subsurface formations and identify potential aquifers. These techniques measure the electrical resistivity of subsurface materials, which varies depending on lithology, rock type, porosity, fluid content, weathering degree, faults, fractures, joints, saturation, and salinity (Singh, 2005; Sinha et al., 2009; Hasan et al., 2021). By integrating resistivity data with geological and hydrogeological information, researchers have been able to optimize the placement of wells, accurately assess groundwater potential, and support the effective and sustainable management of water resources (De Lima and Niwas, 2000; Hubbard and Rubin, 2002; Niwas and De Lima, 2003; Singh, 2005; Soupios et al., 2007; Sinha et al., 2009; Majumdar and Das, 2011; Nwosu et al., 2013; Hasan et al., 2021; Asfahani, 2023). By establishing a useful correlation between electrical resistivity and the limited borehole-based hydraulic parameters, resistivity methods offer the best alternative of the expensive drilling tests for estimating aquifer parameters over large area from shallow to large depths. In this study, we present a novel application of the Controlled Source Audio-Frequency Magnetotelluric (CSAMT) method to develop high-resolution 2D and 3D permeability (k) models at significant depths (up to 1300 meters) within a geologically complex and heterogeneous setting composed of sandstone, granite, and hornstone. Initially, a limited number of boreholes were drilled at strategically selected key locations across the study area. Subsequently, multiple CSAMT survey profiles were conducted, covering the entire region, including the borehole sites. By correlating the resistivity data obtained from the CSAMT surveys with permeability values derived from the borehole data, we established a reliable empirical relationship between resistivity and permeability. This correlation was then applied across the full CSAMT dataset, enabling the generation of 2D and 3D permeability models even in areas lacking direct borehole information. This approach allows for a more comprehensive and cost-effective assessment of deep groundwater resources, significantly reducing the need for extensive and expensive drilling campaigns.”

Comment 3:

Line 139- 140: the statement is too arbitrary; the language can be modified or more evidence is provided.

Response 3:

ORIGINAL STATEMENT:

“No one had ever tried to estimate K using direct or indirect methods in such a heterogeneous context before this work, where a broad diversity of rock types are present at a depth of 1 kilometer”.

REVISED STATEMENT:

“Prior to this study, no attempts had been made to estimate permeability (K) using either direct methods, such as borehole testing, or indirect geophysical approaches in such a geologically heterogeneous setting, characterized by a diverse mixture of sandstone, granite, and hornstone, extending to depths of up to one kilometer.”

Comment 4:

Line 161: in section 2.1 Study area, the author might consider adding more details about the rocks and geology of the study area.

Response 4:

Additional details about the rocks and geological characteristics of the study area have been added, as shown below.

ORIGINAL:

“Intruding rocks from the Indosinian, Caledonian, and Yanshanian eras are among the many geological formations and periods represented in the study region. Other layers from the Paleogene period are also present. The most common types of rock that have been discovered are sandstone, granite, and hornstone. The complex Kaiping concave fault and fold systems were the dominant geological features in the project region, which were developed as a result of magmatic processes and various structures (Qin, 2017). Emergence of joint fissured features symbolizes the various tectono-geological periods, with the local tectonic line corresponding with the faults strike, especially in the northeast orientation (Yang et al., 2021)”.

REVISED:

“The study area exhibits a complex and diverse geological history, characterized by well-defined geometrical relationships among various lithologies. These formations and structural features are the result of multiple tectono-magmatic events spanning several geological periods. Intrusive rocks from the Indosinian (Late Triassic), Caledonian (Silurian–Devonian), and Yanshanian (Jurassic–Cretaceous) orogenies are well-represented, indicating a long sequence of crustal deformation and magmatic activity. These intrusions are primarily composed of granitic bodies, which suggest deep-seated magmatic processes associated with continental collision and subduction zones. In addition to these intrusive phases, sedimentary strata from the Paleogene period are also present, reflecting a later stage of basin development with fluvial and lacustrine depositional environments. Among the most prevalent rock types encountered in the region are sandstone, granite, and hornstone. Sandstone reflects high-energy sedimentary deposition. Granite indicates deep magmatic intrusions likely associated with Yanshanian tectonics. Hornstone (hornfels) results from contact metamorphism caused by magma intruding sedimentary rocks. The structural framework of the region is dominated by the Kaiping concave fault and fold system, a geologically significant and highly deformed zone that reflects multiple deformation episodes (Qin, 2017). These structures were primarily shaped by magmatic intrusions, crustal movements, and regional stress regimes. The presence of extensive jointed and fissured zones throughout the rock mass further supports a history of dynamic tectonic activity. These joints often serve as secondary permeability pathways and are critical in controlling groundwater flow in the fractured rock environment. Importantly, the orientation of these structural features, including faults and joints, is often aligned with northeast-trending tectonic lines, which are consistent with broader regional stress directions (Yang et al., 2021). This relationship among lithologies and structural features plays a critical role in controlling groundwater flow and permeability distribution.”

Comment 5:

Line 204: what does “5-20%” represent for? More specific content is preferred for this sentence.

Response 5:

“The vertical resolution of 5–20% can be assessed by CSAMT when exploring depths ranging from 20 to 1000 meters” explained as in the revised version:

“In CSAMT, the vertical resolution, which refers to the ability to distinguish between adjacent subsurface layers, can typically range between 5% and 20% of the investigation depth from approximately 20 to 1000 meters. At shallower depths (e.g., 20–100 m), vertical resolution is higher (closer to 5%), enabling better differentiation between thin layers. At greater depths (up to 1000 m), resolution may degrade toward the 20% mark due to signal attenuation and broader averaging of resistivity data. This makes CSAMT a valuable tool for identifying significant lithological contrasts, fault zones, and resistivity anomalies related to geological structures”.

Comment 6:

Line 217: in section 2.2.2, why were 6 profiles selected? How did the author determine the locations of the profiles? The author might consider providing more evidence of the site location and data collection in the supporting material.

Response 6:

Further details regarding the selection criteria and rationale for the survey profiles have been provided in the revised manuscript.

ORIGINAL:

“The CSAMT data was acquired using six profiles (1–6) with a 50 meter interval between each station”.

REVISED:

“CSAMT data were acquired along six profiles (profiles 1–6), with a station spacing of 50 meters between each measurement point. The location of 6 CSAMT profiles was chosen based on several factors, including geological targets and objectives, surface geology and mapping data, topography and terrain accessibility, orientation relative to structures, spacing and coverage requirements, resistivity contrast expectations, integration with other data (boreholes), environmental and regulatory constraints, and source-receiver geometry requirements, etc. Carefully selected survey profiles enhanced the ability to resolve critical subsurface features and minimized ambiguities in the geophysical interpretation”.

Additional details are provided in the revised manuscript.

Comment 7:

Line 250-154: The author might consider providing more details of the static correction and the Hanning window spatial filtering method.

Response 7:

Additional details on static correction and Hanning window spatial filtering have been included in the revised manuscript to enhance clarity and support the interpretation of CSAMT data.

ORIGINAL:

“The static corrections were made using a Hanning window spatial filtering method, which involved geological information and curve analysis.”

REVISED:

“Static correction and spatial filtering using a Hanning window are essential preprocessing steps in CSAMT data analysis, aimed at improving data quality and enhancing the reliability of subsurface resistivity models. Static correction addresses the effects of near-surface resistivity inhomogeneities, which can distort electric field measurements and introduce static shifts, vertical displacements in apparent resistivity curves that misrepresent deeper subsurface conditions. This correction typically involves adjusting the measured electric fields by referencing them to a stable or averaged field, effectively removing shallow-layer influences and isolating true subsurface signals. Spatial filtering, on the other hand, is used to mitigate noise introduced by environmental and instrumental sources. Among various filters, the Hanning (Hann) window is commonly applied due to its effectiveness in reducing spectral leakage and smoothing data. When used in spatial filtering, the Hanning window averages measurements across adjacent stations in a weighted manner, preserving spatial trends while suppressing high-frequency noise. This improves the coherence of the dataset and ensures more stable and interpretable inversion results”.

Comment 8:

Figure 1: typos in (b), “uncertainty”; also the words are too small to read.

Response 8:

Figure 1, along with all other figures, has been redrawn and improved for better clarity and presentation. The updated figs are included at the end of the response/comments section in the attached file.

Comment 9:

Figure 7 and Figure 8: a little confused about the legend of the north direction in both figures

Response 9:

The north direction in these figures is correctly oriented, though slightly tilted, to provide a clearer and more informative view of the 3D permeability (k) models. The revised figures are included at the end of the response/comments section in the attached file.

Referee Comments (RC2):

This paper addresses the challenges associated with current groundwater exploration and evaluates the advantages and disadvantages of various methods for measuring hydraulic parameters. The author highlights the application of a novel approach, by using Controlled Source Audio-Frequency Magnetotellurics (CSAMT) method, which is employed to estimate 2D and 3D permeability at depths exceeding 1 km in highly heterogeneous rock environments. The study presents its methods and findings in a well-structured manner, offering insights into deep groundwater exploration.

However, certain assertions appear overly generalized and could benefit from further substantiation. Additionally, more detailed descriptions of the methodologies and the study area would enhance the clarity, reproducibility, and robustness of the research.

Response:

We sincerely thank the anonymous reviewer for their insightful and constructive feedback, which has significantly contributed to enhancing the quality of our work. We have made every effort to revise the manuscript thoroughly in line with the reviewer’s suggestions.

In the revised version, we have expanded the descriptions of both the study area and the methodologies to provide greater clarity and context. Additional explanations and supporting information have also been included to substantiate our findings and address the points raised.

As per the journal’s submission guidelines, we are first submitting our detailed responses to the reviewer’s comments. Following this, we will submit the revised manuscript reflecting all the suggested changes.

Specific comments:

Comment 1:

Line108-119: the author might consider adding a little more evidence of the reason that CSAMT is selected for this study. For example, the author stated that VES method is used to evaluate groundwater resources in a single dimension by a broad of previous studies, but did not illustrate the background about why they did not use other methods, like CSAMT or ERT. Additionally, the author states that there are three main methods, but there are very few examples or introductions about ERT in this paragraph.

Response 1:

The entire paragraph from Lines 88–119 has been revised for improved clarity and structure in the updated manuscript. Both the original and the revised versions of the paragraph are provided below for comparison.

ORIGINAL PARAGRAPH:

“A number of prior groundwater investigations have made use of geophysical techniques (Bentley and Gharibi, 2004; Yadav and Singh, 2007; Fu et al., 2013; Vouillamoz et al., 2014; Robinson et al., 2016; Lin et al., 2018; Kouadio et al., 2020; Abbas et al., 2022; Kouadio et al., 2023; Zhang et al., 2024). A number of studies have shown that geophysical procedures outperform drilling techniques in terms of speed, ease of use, cost, and lack of invasiveness (Hu et al., 2013; Lin et al., 2018; Di et al., 2020; Fusheng et al., 2022; Hasan et al., 2024). Additionally, they are capable of conducting thorough geological evaluations in both the vertical and horizontal planes (Fu et al., 2013; Hasan et al., 2021). These methods are superior to others when it comes to collecting hydrogeological data from various subterranean habitats (Niwas and De Lima, 2003; Wynn et al., 2016; Kouadio et al., 2023). Groundwater studies nowadays often include resistivity surveys. Resistivity methods offer a broader resistivity range compared to other geophysical parameters, which is a major advantage (Bentley and Gharibi, 2004; Camporese et al., 2011; Robinson et al., 2016). The three main methods for measuring resistivity are the controlled source audio-frequency magnetotellurics (CSAMT), vertical electoral soundings (VES), and electrical resistivity tomography (ERT) (Soupios et al., 2007; Di et al., 2020; Zhang et al., 2024). Niwas and De Lima (2003), Soupios et al. (2007), Majumdar and Das (2011), Nwosu et al. (2013), Hasan et al. (2021), and Asfahani (2023) are among the previous groundwater-based geophysical studies that primarily utilized the VES method to evaluate groundwater resources in a single dimension. It is unusual to evaluate aquifer yield at great depths in hard rock terrains using two- and three-dimensional hydraulic properties. Recent studies have demonstrated that CSAMT, which aims to gather extensive subsurface data at very deep depths using 2D/3D evaluations, is the most cost-effective and appropriate geophysical method for researching hard rock (Smith and Booker, 1991; Simpson and Bahr, 2005; Bai et al., 2010; Fu et al., 2013; Hu et al., 2013; Wang et al., 2015; Wynn et al., 2016; Di et al., 2020; Zhang et al., 2021; Kouadio et al., 2023; Hasan et al., 2024). Advantages of CSAMT over other geophysical research methods include its lower cost, its responsiveness to low-resistance rocks, and its ease of usage in challenging topographic circumstances (An et al., 2016; Kouadio et al., 5 115 2020; Zhang et al., 2021). Compared to most geophysical technologies, including ERT, CSAMT's subsurface assessment capabilities are superior due to its depth capacity of up to one kilometer (Zonge and Hughes, 1988; Hasan et al., 2024). When combined with empirically based methodologies, CSAMT becomes an even more powerful tool for studying the incredibly diverse topographical features.”

REVISED PARAGRAPH:

“A number of prior groundwater investigations have made use of geophysical techniques (Bentley and Gharibi, 2004; Yadav and Singh, 2007; Fu et al., 2013; Vouillamoz et al., 2014; Robinson et al., 2016; Lin et al., 2018; Kouadio et al., 2020; Abbas et al., 2022; Kouadio et al., 2023; Zhang et al., 2024). A number of studies have shown that geophysical procedures outperform drilling techniques in terms of speed, ease of use, cost, and lack of invasiveness (Hu et al., 2013; Lin et al., 2018; Di et al., 2020; Fusheng et al., 2022; Hasan et al., 2024). Additionally, they are capable of conducting thorough geological evaluations in both the vertical and horizontal planes (Fu et al., 2013; Hasan et al., 2021). These methods are superior to others when it comes to collecting hydrogeological data from various subterranean habitats (Niwas and De Lima, 2003; Wynn et al., 2016; Kouadio et al., 2023). Groundwater studies nowadays often include resistivity surveys. Resistivity methods offer a broader resistivity range compared to other geophysical parameters, which is a major advantage (Bentley and Gharibi, 2004; Camporese et al., 2011; Robinson et al., 2016). The three main methods for measuring resistivity are the controlled source audio-frequency magnetotellurics (CSAMT), vertical electoral soundings (VES), and electrical resistivity tomography (ERT) (Soupios et al., 2007; Di et al., 2020; Zhang et al., 2024). Niwas and De Lima (2003), Soupios et al. (2007), Majumdar and Das (2011), Nwosu et al. (2013), Hasan et al. (2021), and Asfahani (2023) are among the previous groundwater-based geophysical studies that primarily utilized the VES method to evaluate groundwater resources in a single dimension. In recent decades, electrical resistivity tomography (ERT) has been widely used in hydrogeological studies for 2D and 3D assessment of groundwater resources (Bentley and Gharibi, 2004; Camporese et al., 2011; MLin et al., 2018; Abbas et al., 2022). However, it is unusual to evaluate aquifer yield at great depths in hard rock terrains using two and three dimensional hydraulic properties. Recent studies have demonstrated that CSAMT, which aims to gather extensive subsurface data at very deep depths using 2D/3D evaluations, is the most appropriate geophysical method for researching hard rock (Smith and Booker, 1991; Simpson and Bahr, 2005; Bai et al., 2010; Fu et al., 2013; Hu et al., 2013; Wang et al., 2015; Wynn et al., 2016; Zhang et al., 2021; Kouadio et al., 2023). The selection of resistivity methods in groundwater studies depends on several key factors, including survey objectives, depth of investigation, resolution, geological complexity, logistical constraints, cost and accessibility, electrical conductivity contrast, and field conditions, etc (Di et al., 2020; Hasan et al., 2024). VES is more suitable for shallow depths (< 200 m) for 1D resistivity imaging, has limited lateral resolution, is useful for layered aquifer characterization, works well in horizontally stratified formations, requires minimal equipment and is quick to deploy, is the most economical for small-scale studies, and may face limitations in highly resistive or conductive terrains (Soupios et al., 2007; Nwosu et al., 2013; Hasan et al., 2021). ERT offers better resolution for both shallow and intermediate depths (up to ~300 m) with 2D/3D imaging, provides high-resolution subsurface imaging, is ideal for detecting fractures/faults and heterogeneous aquifers, is better for complex geology (e.g., fractured zones, karst systems), is ideal for detailed aquifer geometry and contamination studies, needs more field effort and electrode spacing adjustments, and may face limitations in highly resistive or conductive terrains like VES (Camporese et al., 2011; MLin et al., 2018; Abbas et al., 2022). CSAMT is effective for deeper investigations (hundreds to thousands of meters) due to its low-frequency signal penetration, provides 2D/3D imaging over big area at large depths, has lower resolution than ERT but excels in deep structural mapping, is preferred for deep-seated structures like basement aquifers or geothermal systems, is used for regional groundwater exploration, demands specialized equipment and is more time-consuming compared with VES and ERT, is relatively more expensive than VES and ERT due to advanced instrumentation and processing, all resistivity methods rely on resistivity contrasts but CSAMT is more sensitive to deep conductive zones, and performs better in areas with cultural noise (e.g., urban settings) due to controlled signal sources (Zonge and Hughes, 1988; An et al., 2016; Kouadio et al., 2020; Zhang et al., 2021). When combined with empirically based methodologies, CSAMT becomes an even more powerful tool for studying the incredibly diverse topographical features at large depths (Hasan et al., 2024). So based on above factors, CSAMT was the most suitable method for this study.”

Comment 2:

Line 120- 138: the author might consider reorganizing this paragraph to make the significance of the resistivity method stand out.

Response 2:

The paragraph previously located in Lines 120–138 has been reorganized for improved clarity in the revised manuscript. For reference, both the original and the revised versions of the paragraph are provided below.

ORIGINAL PARAGRAPH:

“Several factors, such as the type of rock, fault, weathering degree, fluid content, permeability, pore-spacing, fracture, lithology, saturation, and joints, as well as the same structural heterogeneities, determine the geophysical and aquifer characteristics (Singh, 2005; Sinha et al., 2009; Hasan et al., 2021). Several prior studies utilized geophysical parameters in conjunction with hydraulic data or lithological logs to characterize underlying rock mass units hydrogeologically (De Lima and Niwas, 2000; Hubbard and Rubin, 2002; Niwas and De Lima, 2003; Singh, 2005; Soupios et al., 2007; Sinha et al., 2009; Majumdar and Das, 2011; Nwosu et al., 2013; Hasan et al., 2021; Asfahani, 2023). Resistivity methods provide an alternate option for aquifer parameter estimation by creating a beneficial relationship between electrical resistivity and the aquifer parameters (obtained from drilling tests). An innovative aspect of this work is its use of non-invasive geophysical techniques to create two- and three-dimensional k models in a diverse environment with a variety of rock types and significant depths. The planned study will necessitate the boring of a handful of boreholes at key spots all around the project site. A more trustworthy CSAMT study will allow us to evaluate the extensive research area. Then, by directly connecting geophysical and borehole data, k can be established for the entire researched site, even without drilling tests. Two- and three-dimensional k models are generated by applying the resulting equations to the full study area. This approach would reduce the need for costly boreholes to obtain a thorough and complete evaluation of subsurface hydrogeological conditions.”

REVISED PARAGRAPH:

“Resistivity methods are highly significant in groundwater studies due to their ability to characterize subsurface formations and identify potential aquifers. These techniques measure the electrical resistivity of subsurface materials, which varies depending on lithology, rock type, porosity, fluid content, weathering degree, faults, fractures, joints, saturation, and salinity (Singh, 2005; Sinha et al., 2009; Hasan et al., 2021). By integrating resistivity data with geological and hydrogeological information, researchers have been able to optimize the placement of wells, accurately assess groundwater potential, and support the effective and sustainable management of water resources (De Lima and Niwas, 2000; Hubbard and Rubin, 2002; Niwas and De Lima, 2003; Singh, 2005; Soupios et al., 2007; Sinha et al., 2009; Majumdar and Das, 2011; Nwosu et al., 2013; Hasan et al., 2021; Asfahani, 2023). By establishing a useful correlation between electrical resistivity and the limited borehole-based hydraulic parameters, resistivity methods offer the best alternative of the expensive drilling tests for estimating aquifer parameters over large area from shallow to large depths. In this study, we present a novel application of the Controlled Source Audio-Frequency Magnetotelluric (CSAMT) method to develop high-resolution 2D and 3D permeability (k) models at significant depths (up to 1300 meters) within a geologically complex and heterogeneous setting composed of sandstone, granite, and hornstone. Initially, a limited number of boreholes were drilled at strategically selected key locations across the study area. Subsequently, multiple CSAMT survey profiles were conducted, covering the entire region, including the borehole sites. By correlating the resistivity data obtained from the CSAMT surveys with permeability values derived from the borehole data, we established a reliable empirical relationship between resistivity and permeability. This correlation was then applied across the full CSAMT dataset, enabling the generation of 2D and 3D permeability models even in areas lacking direct borehole information. This approach allows for a more comprehensive and cost-effective assessment of deep groundwater resources, significantly reducing the need for extensive and expensive drilling campaigns.”

Comment 3:

Line 139- 140: the statement is too arbitrary; the language can be modified or more evidence is provided.

Response 3:

ORIGINAL STATEMENT:

“No one had ever tried to estimate K using direct or indirect methods in such a heterogeneous context before this work, where a broad diversity of rock types are present at a depth of 1 kilometer”.

REVISED STATEMENT:

“Prior to this study, no attempts had been made to estimate permeability (K) using either direct methods, such as borehole testing, or indirect geophysical approaches in such a geologically heterogeneous setting, characterized by a diverse mixture of sandstone, granite, and hornstone, extending to depths of up to one kilometer.”

Comment 4:

Line 161: in section 2.1 Study area, the author might consider adding more details about the rocks and geology of the study area.

Response 4:

Additional details about the rocks and geological characteristics of the study area have been added, as shown below.

ORIGINAL:

“Intruding rocks from the Indosinian, Caledonian, and Yanshanian eras are among the many geological formations and periods represented in the study region. Other layers from the Paleogene period are also present. The most common types of rock that have been discovered are sandstone, granite, and hornstone. The complex Kaiping concave fault and fold systems were the dominant geological features in the project region, which were developed as a result of magmatic processes and various structures (Qin, 2017). Emergence of joint fissured features symbolizes the various tectono-geological periods, with the local tectonic line corresponding with the faults strike, especially in the northeast orientation (Yang et al., 2021)”.

REVISED:

“The study area exhibits a complex and diverse geological history, characterized by well-defined geometrical relationships among various lithologies. These formations and structural features are the result of multiple tectono-magmatic events spanning several geological periods. Intrusive rocks from the Indosinian (Late Triassic), Caledonian (Silurian–Devonian), and Yanshanian (Jurassic–Cretaceous) orogenies are well-represented, indicating a long sequence of crustal deformation and magmatic activity. These intrusions are primarily composed of granitic bodies, which suggest deep-seated magmatic processes associated with continental collision and subduction zones. In addition to these intrusive phases, sedimentary strata from the Paleogene period are also present, reflecting a later stage of basin development with fluvial and lacustrine depositional environments. Among the most prevalent rock types encountered in the region are sandstone, granite, and hornstone. Sandstone reflects high-energy sedimentary deposition. Granite indicates deep magmatic intrusions likely associated with Yanshanian tectonics. Hornstone (hornfels) results from contact metamorphism caused by magma intruding sedimentary rocks. The structural framework of the region is dominated by the Kaiping concave fault and fold system, a geologically significant and highly deformed zone that reflects multiple deformation episodes (Qin, 2017). These structures were primarily shaped by magmatic intrusions, crustal movements, and regional stress regimes. The presence of extensive jointed and fissured zones throughout the rock mass further supports a history of dynamic tectonic activity. These joints often serve as secondary permeability pathways and are critical in controlling groundwater flow in the fractured rock environment. Importantly, the orientation of these structural features, including faults and joints, is often aligned with northeast-trending tectonic lines, which are consistent with broader regional stress directions (Yang et al., 2021). This relationship among lithologies and structural features plays a critical role in controlling groundwater flow and permeability distribution.”

Comment 5:

Line 204: what does “5-20%” represent for? More specific content is preferred for this sentence.

Response 5:

“The vertical resolution of 5–20% can be assessed by CSAMT when exploring depths ranging from 20 to 1000 meters” explained as in the revised version:

“In CSAMT, the vertical resolution, which refers to the ability to distinguish between adjacent subsurface layers, can typically range between 5% and 20% of the investigation depth from approximately 20 to 1000 meters. At shallower depths (e.g., 20–100 m), vertical resolution is higher (closer to 5%), enabling better differentiation between thin layers. At greater depths (up to 1000 m), resolution may degrade toward the 20% mark due to signal attenuation and broader averaging of resistivity data. This makes CSAMT a valuable tool for identifying significant lithological contrasts, fault zones, and resistivity anomalies related to geological structures”.

Comment 6:

Line 217: in section 2.2.2, why were 6 profiles selected? How did the author determine the locations of the profiles? The author might consider providing more evidence of the site location and data collection in the supporting material.

Response 6:

Further details regarding the selection criteria and rationale for the survey profiles have been provided in the revised manuscript.

ORIGINAL:

“The CSAMT data was acquired using six profiles (1–6) with a 50 meter interval between each station”.

REVISED:

“CSAMT data were acquired along six profiles (profiles 1–6), with a station spacing of 50 meters between each measurement point. The location of 6 CSAMT profiles was chosen based on several factors, including geological targets and objectives, surface geology and mapping data, topography and terrain accessibility, orientation relative to structures, spacing and coverage requirements, resistivity contrast expectations, integration with other data (boreholes), environmental and regulatory constraints, and source-receiver geometry requirements, etc. Carefully selected survey profiles enhanced the ability to resolve critical subsurface features and minimized ambiguities in the geophysical interpretation”.

Additional details are provided in the revised manuscript.

Comment 7:

Line 250-154: The author might consider providing more details of the static correction and the Hanning window spatial filtering method.

Response 7:

Additional details on static correction and Hanning window spatial filtering have been included in the revised manuscript to enhance clarity and support the interpretation of CSAMT data.

ORIGINAL:

“The static corrections were made using a Hanning window spatial filtering method, which involved geological information and curve analysis.”

REVISED:

“Static correction and spatial filtering using a Hanning window are essential preprocessing steps in CSAMT data analysis, aimed at improving data quality and enhancing the reliability of subsurface resistivity models. Static correction addresses the effects of near-surface resistivity inhomogeneities, which can distort electric field measurements and introduce static shifts, vertical displacements in apparent resistivity curves that misrepresent deeper subsurface conditions. This correction typically involves adjusting the measured electric fields by referencing them to a stable or averaged field, effectively removing shallow-layer influences and isolating true subsurface signals. Spatial filtering, on the other hand, is used to mitigate noise introduced by environmental and instrumental sources. Among various filters, the Hanning (Hann) window is commonly applied due to its effectiveness in reducing spectral leakage and smoothing data. When used in spatial filtering, the Hanning window averages measurements across adjacent stations in a weighted manner, preserving spatial trends while suppressing high-frequency noise. This improves the coherence of the dataset and ensures more stable and interpretable inversion results”.

Comment 8:

Figure 1: typos in (b), “uncertainty”; also the words are too small to read.

Response 8:

Figure 1, along with all other figures, has been redrawn and improved for better clarity and presentation. The updated figs are included at the end of the response/comments section in the attached file.

Comment 9:

Figure 7 and Figure 8: a little confused about the legend of the north direction in both figures

Response 9:

The north direction in these figures is correctly oriented, though slightly tilted, to provide a clearer and more informative view of the 3D permeability (k) models. The revised figures are included at the end of the response/comments section in the attached file.