Multi-Sensor Remote Sensing Evidence of Connection between Deep Aquifer Recharge in California's Central Valley and Sierra Nevada Snowmelt

Werth, Susanna; Shirzaei, Manoochehr; Carlson, Grace; Bürgmann, Roland

doi:10.5194/egusphere-2026-3061

Preprints

https://doi.org/10.5194/egusphere-2026-3061

Preprints

11 Jun 2026

| 11 Jun 2026

Multi-Sensor Remote Sensing Evidence of Connection between Deep Aquifer Recharge in California's Central Valley and Sierra Nevada Snowmelt

Susanna Werth, Manoochehr Shirzaei, Grace Carlson, and Roland Bürgmann

Abstract. California’s arid Central Valley (CV) relies heavily on groundwater extracted from deep aquifers (>50 m) and imported surface water to sustain approximately one-quarter of the United States’ food production. While the recharge to deep aquifers in CV is hypothesized to be influenced by hydrologic processes in the adjacent Sierra Nevada, direct observational evidence at regional scales remains limited.

Here, we present an integrated analysis of multi-decadal geodetic remote sensing, hydrologic, and climate datasets, including groundwater levels, GRACE-derived water storage, GNSS and InSAR deformation, and snowmelt observations, to investigate the spatiotemporal relationships between mountain hydrology and deep aquifer dynamics in the CV. A consistent sequence of seasonal signals emerges across independent datasets: peak groundwater levels in deep CV aquifers occur approximately one month after peak water availability from Sierra Nevada snowmelt, while peak groundwater storage inferred from GRACE lags groundwater levels by approximately three months.

These phase delays indicate that groundwater systems respond to mountain hydrologic forcing through temporally distinct processes, with pressure signals propagating more rapidly than changes in bulk water storage. A simplified first-order diffusion analysis shows that the observed lag between snowmelt and the groundwater-level response is feasible, given plausible pressure propagation timescales in fractured mountain bedrock.

Together, these results provide observational evidence consistent with a hydraulic connection between high-mountain aquifers and deep basin aquifers via mountain block recharge (MBR). Our findings highlight the importance of incorporating mountain-driven recharge processes and pressure dynamics into regional hydroclimate models and groundwater management strategies, particularly in snowmelt-dominated and water-stressed regions.

Received: 28 May 2026 – Discussion started: 11 Jun 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 20791 KB)

Supplement (16114 KB)

Download & links

Preprint (20791 KB)
Metadata XML
Supplement (16114 KB)
BibTeX
EndNote

Susanna Werth, Manoochehr Shirzaei, Grace Carlson, and Roland Bürgmann

Status: final response (author comments only)

RC1:
'Comment on egusphere-2026-3061', Anonymous Referee #1, 09 Jul 2026

GENERAL COMMENT

The paper addresses relevant scientific questions that fall well within the scope of the journal. The manuscript lays the groundwork for further studies and investigations aimed at characterizing important recharge mechanisms in valley aquifers and their interactions with adjacent mountain systems. The results presented appear sufficient to support the interpretations and conclusions drawn by the authors, who also provide appropriate credit to previous work while clearly highlighting the novel contributions of the present study. In addition, the title accurately reflects the content of the paper, and the abstract provides a concise and comprehensive summary of both the analyses performed and the main findings.

A few references appear to be missing throughout the manuscript: these are indicated in the “Technical corrections” section below. I would also like to comment on the substantial amount of the supplementary material provided. In my opinion, including such a large amount of additional material may make it more difficult for readers to identify the key results and conclusions of the study. I would therefore strongly encourage the authors to consider reducing the supplementary material by removing sections that are not essential for understanding the analyses presented in the manuscript.

SPECIFIC COMMENTS

1. It seems to me that the statement in lines 38-40 (“A direct consequence [...] (Galloway et al., 2000; Herrera-García et al., 2021).”) would benefit from additional context. It appears that a word may be missing (e.g., “deep confined aquifer exploitation” or “deep confined aquifer depletion”), which could help clarify why this statement is included at this point in the manuscript. If the discussion concerns both aquifer recharge and depletion, the term “replenish” in line 40 might be better replaced by wording referring to both filling and depletion processes. Otherwise, if the discussion concerns recharge only, the sentence in line 38-40 might be more appropriately relocated elsewhere in the manuscript.

2. Lines 94-95. The connection between the low groundwater flow velocity and the expectation that deep lowland aquifers are recharged by lateral flows originating at higher elevations is not entirely clear to me. Could the authors please provide some clarification on this point?

3. Line 158. I am not entirely sure that the term “data-driven” (also al lines 164 and 640) is appropriate in this context, although it may simply reflect my limited familiarity with the method. Could the authors please provide some clarification on this point?

4. Lines 158-159. For greater clarity and completeness, the authors may whish to briefly justify here the choice of the methods adopted in the analyses presented in the manuscript.

5. Line 181. The authors may consider including a brief explanation of what is meant by mascon cells. Readers who are not familiar with the topic may otherwise find this part difficult to follow.

6. Lines 193-198. If I understood correctly, this section refers to SnS rather than SoS, as currently written in lines 193 and 197. SoS is instead discussed in lines 186-191. If this is the case, I suggest correcting the terminology accordingly.

7. Lines 204-207. The purpose and relevance of the analysis described in these lines within the context of the present study are not entirely clear to me. The authors may wish to provide additional context or motivation.

8. Lines 219-220. For greater clarity and transparency, I would suggest explicitly stating the additional selection criteria based on the distance from pumping wells and on the historical stability of water levels.

9. Lines 222-223. On what basis was this method selected for outlier detection and removal? Is this a standard procedure in similar studies?

10. Lines 226-227. Is the requirement for time series spanning at least 3 years and containing less than 3 months of missing data an additional criterion for selecting the wells used in the subsequent analyses? This is not entirely clear from the current wording.

11. Line 233. The meaning of the term “unique” in this context is not entirely clear to me.

12. Lines 241-242. I am not entirely clear on why an increase in water mass would lead to subsidence, whereas a decrease in mass would result in surface uplift. I would intuitively expect the opposite, although this may simply reflect my limited familiarity with the topic. In any case, I would suggest including a brief explanation in the manuscript to facilitate understanding for readers who may not be fully familiar with the underlying processes.

13. Similarly, I would suggest expanding slightly on the explanation of the poroelastic deformation process described here.

14. Line 232. If the quantities hp2 and hp1 mentioned here correspond to h2 and h1 in Figure 1, the authors may consider adopting a consistent notation throughout the manuscript to facilitate the comprehension of the process described.

15. Lines 333-334. To what extent are the assumptions mentioned here realistic and applicable to the considered study area?

16. Line 480. The rationale for selecting k=0.3 m2/s is provided later in the manuscript (line 575). However, I would suggest either moving this explanation to this earlier point or briefly adding it here when the value is first introduced, so that the choice is immediately justified.

TECHNICAL CORRECTIONS

17. Lines 58-59. In my opinion, “water for later use” reads more naturally than “water for use later”. Please consider changing the text.

18. Line 94. There is a repetition of “However”, which is also used in line 92. I would also suggest starting a new paragraph at this point in order to give greater prominence to the subsequent discussion of the two recharge mechanisms (MFR and MBR).

19. Line 107. If the “recent study” referred to here is the one cited in the following line (Armengol et al., 2024), it would be preferable to state this explicitly here. Otherwise, please include the appropriate reference.

20. Lines 136-138. For completeness, the authors may consider to include references for GRACE, GNSS and InSAR.

21. Line 156. It may be preferable to use a different term instead of “feasible”, for example “plausible” or “realistic”.

22. Line 156. After point (3), please correct the singular/plural agreement by using either “have a […] signature” or “have […] signatures”.

23. Line 179. If possible, please include a reference for the RL06 Level-3 product from NASA’s Jet Propulsion Laboratory.

24. Line 203. The abbreviation “SN” for Sierra Nevada is not introduced previously in the text and only appears later in the caption of Figure 6. The authors may consider either to define it when first used or to remove it.

25. Line 252. The expression “the most severe effects” may be clearer than “the effects more severe”; the authors may wish to consider this modification.

26. Line 257. If possible, please include a reference for the GipsyX software.

27. Linnes 276-277. The acronym LOS (line of sight) is already defined in line 266 and does not need to be introduced again here.

28. Lines 285. In most references containing two authors, the names are separated by “&”, whereas here they are separated by “and”. Please verify consistency throughout the manuscript.

29. Please verify that the figures included in the supplementary material appear in the same order in which they are first referenced in the main manuscript.

30. Figure 5. I would suggest swapping the positions of the maps currently shown in Figures 5b and 5c, so that the panels are ordered row-wise rather than column-wise, consistently with the other figures in the manuscript.

31. Please aim to maintain consistency throughout the manuscript in the way multiple figures are cited within the text. For example, in line 468 Figures 6e and 6g are separated by a conjunction (“Fig. 6e and 6g”), whereas in line 471 Figures S13b and S13c are separated by a comma without repeating the figure number (“Fig. S13b,c”). Similarly, line 491 uses “Figs.” Instead of “Fig.” before a list of two figures.

32. Line 482. Please add a space between “100” and “m”, so that the text reads “100 m” rather than “100m”.

33. Line 624. The suffix “st” in “21st” should be superscripted.

34. Line 651. There currently appear to be unnecessary parentheses around the two references cited.

Citation: https://doi.org/10.5194/egusphere-2026-3061-RC1
- AC1:
  'Reply on RC1', Susanna Werth, 24 Jul 2026
  Thank you very much for your supportive evaluation and for taking the time to provide such a detailed and constructive review. We appreciate your thoughtful comments and will carefully address each point in a revised version of the manuscript. Below, we provide responses to the questions and suggestions raised in the review.
  We also thank the reviewer for identifying missing references; these will be added in the revised manuscript.
  Thank you for the suggestion regarding the supplementary material. We agree that it should remain focused and easy to navigate. The included material either provides technical details needed for transparency and reproducibility or presents supporting data for results shown in the main figures. To improve accessibility, we will add a detailed table of contents at the beginning of the Supplement that identifies each section and figure and briefly explains its purpose and relationship to the main manuscript. We will also review the material for opportunities to consolidate redundant content.
  Responses to specific comments:
  We agree that the current wording is incomplete and may obscure the intended link between deep confined aquifer exploitation, groundwater depletion, and land subsidence. We will revise the passage to make this connection explicit and clarify that the discussion concerns both aquifer depletion and replenishment.
  
  Thank you for pointing this out. We will revise the passage to clarify that, in layered basin sediments, vertical hydraulic conductivity is commonly much lower than horizontal conductivity, which limits rapid vertical recharge to deep confined aquifers. By contrast, lateral hydraulic connectivity along permeable layers extending toward the valley margins and mountain front, together with fractured bedrock pathways, may allow higher-elevation mountain aquifers to act as recharge sources for deep basin aquifers (the latter is illustrated in Fig. 2). In the Central Valley, the Sierra Nevada may therefore contribute to deep aquifer recharge through such lateral pathways.
  
  Thank you for raising this point. We use the term “data-driven” because the inferred relationships are derived from statistical and time-frequency analyses of multiple observational datasets rather than prescribed by a process-based model. We will clarify this meaning in the manuscript and reconsider the terminology where appropriate.
  
  We agree that the methodological rationale should be stated more explicitly and will add this explanation to the revised manuscript. The selected datasets provide complementary observations of groundwater mass, hydraulic head, and surface deformation, with each reflecting a different physical component of the coupled system. We focus specifically on the seasonal signal because it represents a recurring pattern of groundwater recharge and discharge captured across multiple sensors. Wavelet analysis is used to isolate these non-stationary seasonal signals and compare their relative timing, allowing us to identify consistent temporal relationships among the observations. By combining these complementary datasets and their temporal dynamics, we aim to obtain a more complete picture of the coupled mountain–valley groundwater system than would be possible from any single static observation.
  
  We agree that this term should be defined for readers unfamiliar with GRACE products. We will add the following explanation to the revised manuscript: “A GRACE mascon cell is a predefined geographic area for which time-varying mass change is estimated from satellite observations and represented as a spatially uniform equivalent water height.”
  
  We agree that this section refers to snow storage (SnS), not soil storage (SoS). We will correct the terminology in lines 193 and 197 accordingly.
  
  We use the SNODAS precipitation and snowmelt variables to characterize the timing of water input from the Sierra Nevada and to compare these fluxes with the seasonal groundwater response in the Central Valley. We will add this motivation to this passage of the revised manuscript.
  
  We agree that the current wording is ambiguous. We relied on the well-use classifications provided by the USGS and CDWR, retaining wells categorized as observation wells and excluding those categorized as irrigation wells. The references to distance from pumping wells and historical water-level stability were not additional quantitative criteria applied independently in our analysis. We will verify the basis of the USGS and CDWR classifications in the accompanying data documentation and revise the passage to make this distinction explicit.
  
  We clarify that the intended threshold was three standard deviations, not 3.5. Assuming an approximately normal distribution, a threshold of 3 times the standard deviation retains about 99.7% of the observations. This was selected as a conservative quality-control filter to remove isolated extreme values while preserving the seasonal variability of the detrended records. We will correct the threshold and clarify that this is a pragmatic criterion rather than a field-specific standard.
  
  The requirement that records span at least three years and contain less than three months of missing data was an additional selection criterion for the subsequent analyses. A minimum of three years of observation length and the absence of longer gaps are necessary to support the wavelet analysis and obtain a stable estimate of seasonal variability. We will state this explicitly in the revised manuscript.
  
  The term unique refers to distinct well locations, because several sites contain multiple nested monitoring wells or sensors at different depths. We will revise the wording to make this clear.
  
  Thank you for pointing out that this explanation may not be sufficiently intuitive. An increase in total water mass increases the elastic load on Earth’s crust and causes subsidence, whereas a decrease in mass reduces the load and causes uplift. This elastic loading response overlaps spatially with poroelastic aquifer deformation, which has the opposite sign for a given groundwater change and can be substantially larger in magnitude. Because the cited passage introduces both mechanisms quite briefly and in close succession, it may be confusing for readers who are more familiar with either deformation process. We will reformulate and expand the explanation of the deformation processes to distinguish the two contrasting mechanisms more clearly.
  
  Thank you for the comment. We will expand the explanation of this deformation process. We will clarify that the poroelastic deformation process results from declining water pressure in aquifers, which increases effective stress on the aquifer skeleton and causes compaction of underground layers and surface subsidence, whereas rising pressure reduces effective stress and can produce elastic expansion and uplift.
  
  We believe this comment refers to line 323 rather than line 223. The notation refers to the conceptual hydraulic heads shown in Fig. 2b and is not intended as a separate quantitative formulation. We will revise the notation and accompanying text to improve consistency and clarity.
  
  For the purpose of a first-order assessment, which is our objective in the study, the assumptions are appropriate for the study area. Most importantly, they allow us to test whether pressure propagation over the observed timescale is physically feasible under plausible hydraulic conditions. However, the simplified model cannot identify the exact recharge pathways or determine their precise travel times or storage times within the heterogeneous three-dimensional mountain–valley system. We will clarify this distinction more explicitly in the relevant Methods section. We will further emphasize in the Discussion the need for detailed hydrogeological characterization and process-based modeling to resolve these questions.
  
  We agree that the rationale for selecting k=0.3m²/s should be provided when the value is first introduced. We will give an explanation consistent with the later discussion at this earlier point in the revised manuscript.
  
  Responses to technical comments:
  17-19, 21, 22, 24, 25, 27-29, 31-34)
  
  Noted; these points will be corrected in the revised manuscript.
  20, 23, 26)
  
  Thank you for the suggestion. We will add the requested references. We will further check all references to make sure they are complete in the manuscript.
  30) We agree with this suggestion and will swap the positions of panels 5b and 5c so that the figure is ordered row-wise, consistent with the other figures in the manuscript.
  
  Citation: https://doi.org/10.5194/egusphere-2026-3061-AC1
RC2:
'Comment on egusphere-2026-3061', Anonymous Referee #2, 08 Aug 2026

Establishing the relationship between mountain snow and deep aquifer in view of groundwater sustainability is an important topic and the authors do a very good job of establishing the relationship using multiple methods and datasets (in situ and remote sensing). The work also serves as a foundation for future work aimed at exploring MBR fluxes in same and other similar regions. The authors also appropriately acknowledge the uncertainties and limitations of the data set and methods adopted. The analysis is robust and corroborated using different data. However, the analysis is based on correlation rather than causative analysis that potentially weaken the use of the results in a meaningful way as an input for construction of hydrological models.

Citation: https://doi.org/10.5194/egusphere-2026-3061-RC2
- AC2: 'Reply on RC2', Susanna Werth, 17 Aug 2026
  
  Thank you for taking the time to review our manuscript and for your constructive assessment. We appreciate the reviewer’s recognition that the study is robust, well supported by multiple independent datasets and methods, and provides an important foundation for future work on mountain-block recharge. We also appreciate the suggestions for strengthening the interpretation and clarifying the scope of the conclusions.
  We agree that the relationships identified here should not be interpreted as proof of causality in the strict statistical or experimental sense. As emphasized in the manuscript, our objective is a first-order, multi-sensor assessment of the temporal relationships among mountain water availability, groundwater response, and geodetic observations. The consistency of the observed temporal relationships across multiple independent datasets, together with their agreement with the expected physical sequence linking mountain snow accumulation, snowmelt, recharge, and deep-aquifer response, provides convergent evidence for a hydrologic connection. These observations therefore offer empirical constraints for future physically based groundwater models designed to quantify mountain-block recharge fluxes and explicitly test the mechanisms inferred here. We will revise the Methods and Discussion to make the associative nature and intended scope of the analysis more explicit, and we will expand the Discussion of the geological and geophysical constraints that would be required to test the inferred pathways more directly.
  These observational constraints are valuable even without spatially explicit pathway validation because they provide independent temporal relationships and physically testable hypotheses that future process-based models should be able to reproduce; we do not intend the present results themselves to directly parameterize or validate such models. At the same time, our analysis is not designed to identify or validate the specific three-dimensional pathways through which this connection occurs. Doing so would require spatially explicit hydrogeological characterization and process-based groundwater modeling incorporating geological structure, hydraulic properties, and boundary conditions, which represents a substantially different analysis from the first-order assessment conducted here.
  We would also like to emphasize that the focus on timing is intentional and represents a central contribution of the study rather than a limitation of the analysis. The multi-sensor timing relationships provide independent constraints on the sequence and characteristic timescales of the mountain–aquifer response that are not readily obtained from spatial geological or geophysical mapping alone. Spatial characterization of potential pathways, such as that provided by Kang et al. (2021, 2025), is highly complementary to this temporal perspective, but addressing it in comparable detail would constitute a distinct extension of the present study and shift the analysis from a basin-scale temporal assessment toward a more spatially resolved hydrogeophysical investigation.
  We agree with the reviewer that comparison with existing geological and geophysical models, including information on permeable sedimentary units, buried channels, and structural controls, represents an important next step for testing the mechanisms suggested by our observations. Incorporating additional geophysical observations, such as the airborne electromagnetic (AEM) data used by Kang et al. (2025), could provide valuable spatial constraints, but would substantially expand the scope of the present study and is therefore better suited to a follow-up investigation. In addition, these data do not provide a direct spatial validation of the deep mountain-block recharge pathways investigated here.
  The AEM analysis specifically characterizes sedimentary pathways between the land surface and the water table, reported at approximately 30–100 m depth, and is therefore primarily relevant to shallow recharge from the land surface rather than to continuous hydraulic pathways connecting the Sierra Nevada mountain block with the deeper confined aquifer system. It therefore does not resolve continuous hydraulic pathways from the Sierra Nevada mountain block into the deeper Central Valley aquifer system considered in our study. We will expand the Discussion to make this distinction clearer and to identify integration with three-dimensional hydrogeological and geophysical data and groundwater-flow models as a priority for future work. We will also revise the Methods and Discussion where necessary to ensure that the associative nature and intended scope of the present analysis are clearly stated.
  We would also like to clarify that we do not attribute the GRACE-derived groundwater-storage signal specifically to the deep aquifer system. As stated in the manuscript, after accounting for the other major water-storage components, GRACE-derived groundwater storage represents an integrated signal of groundwater-storage changes across both shallow and deep aquifers and cannot resolve storage changes by depth. The depth-specific evidence for the deep-aquifer response in our analysis is instead provided primarily by the groundwater-level observations, while GRACE provides an independent basin-scale constraint on groundwater mass variability. We will review the relevant text to ensure that this distinction is sufficiently clear.
  All datasets used in this analysis, including GRACE-derived groundwater storage, GNSS vertical displacement, groundwater levels, and InSAR-derived deformation, are expressed as anomalies relative to a common baseline period prior to comparison. We recognize that this was not stated with sufficient clarity in the original submission, and we will revise the Methods to state explicitly that all comparisons are conducted on anomaly, rather than observed, values, so that the basis for cross-dataset comparability is unambiguous.
  We also acknowledge the limitation raised regarding the depth sensitivity of the InSAR-derived deformation signal. Surface deformation observed by InSAR reflects the integrated surface response to deformation occurring throughout the aquifer system and cannot, on its own, be assigned to a discrete depth interval. In our analysis, InSAR observations are used specifically to assess the timing and phase relationship of the deformation response relative to the other datasets, rather than to attribute deformation magnitude to a particular well-screen depth. We note, however, that our analysis in Supplementary Text S1 indicates that the deformation contribution from the unconfined aquifer is negligible. We therefore interpret the InSAR deformation signal as being dominated by deformation within the confined and semi-confined aquifer system, while recognizing that InSAR itself does not resolve the depth of the deformation source. We will revise the Discussion to state this constraint explicitly and to clarify that our use of InSAR is limited to this timing-based role.
  Regarding the role of the Corcoran Clay, we recognize that the manuscript could be read as treating the unit as a uniformly continuous confining layer while separately acknowledging the presence of clay lenses elsewhere in the system, and we will revise the text to resolve this more clearly. Our interpretation is that the Corcoran Clay behaves as a regionally extensive but not fully continuous confining layer: where it remains laterally continuous, it supports the separation between shallow and deep system responses that our timing analysis relies on, while locally thinner sections or breaches associated with clay lenses would be expected to shorten the pressure-diffusion pathway between the two systems rather than introduce a competing or unrelated signal. We will note this explicitly as a source of local heterogeneity that cannot be resolved by the present first-order analysis and identify it as a further target for the geological comparison work proposed above.
  Finally, with respect to mountain-front recharge, we agree that MFR represents an important and physically distinct contribution to total mountain-system recharge, and that MFR and MBR together constitute the more complete regional recharge budget. We restricted the present analysis to MBR because our hypothesis specifically concerns hydraulic connection through the mountain block, for which the observed timing can be evaluated using a first-order pressure-diffusion framework. This differs from MFR, which is governed by episodic infiltration from mountain-front stream channels and would require a differently designed analysis centered on stream-aquifer exchange rather than the basin-scale geodetic framework applied here. Combining the two mechanisms within the current analytical framework risks conflating two hydrologically distinct recharge processes. We will add a paragraph to the Discussion stating this scope decision explicitly, acknowledging MFR's contribution to the total recharge budget, and identifying MFR quantification and eventual integration into a combined MSR assessment as a priority direction for future work.
  Kang, S., Knight, R., Greene, T., Buck, C., & Fogg, G. (2021). Exploring the model space of airborne electromagnetic data to delineate large-scale structure and heterogeneity within an aquifer system. Water Resources Research, 57, e2021WR029699. https://doi.org/10.1029/2021WR029699
  Kang, S., Goebel, M., & Knight, R. (2025). Harnessing the power of geophysical imaging to recharge California's groundwater. Earth and Space Science, 12, e2024EA003958. https://doi.org/10.1029/2024EA003958
  
  Citation: https://doi.org/10.5194/egusphere-2026-3061-AC2
CC1:
'Comment on egusphere-2026-3061', Matias Taucare, 11 Aug 2026

Publisher’s note: the content of this comment was removed on 12 August 2026 since the comment was posted by mistake.

Citation: https://doi.org/10.5194/egusphere-2026-3061-CC1
- AC3: 'Reply on CC1', Susanna Werth, 17 Aug 2026
  
  This comment is the same as that of RC3, which we will respond shortly under the RC3 entry.
  
  Citation: https://doi.org/10.5194/egusphere-2026-3061-AC3
RC3: 'Comment on egusphere-2026-3061', Matias Taucare, 12 Aug 2026

Thank you for the opportunity to review the manuscript entitled "Multi-Sensor Remote Sensing Evidence of Connection between Deep Aquifer Recharge in California's Central Valley and Sierra Nevada Snowmelt" by Werth et al. The manuscript addresses an important topic in mountain hydrogeology: the hydraulic connection between high-elevation mountain systems and adjacent sedimentary aquifers. In particular, the authors investigate whether hydrological forcing associated with Sierra Nevada snowmelt can be detected in the groundwater dynamics of the Central Valley using an integrated set of hydrological and geodetic observations. The topic is timely, the multi-sensor approach is potentially valuable, and I consider the manuscript suitable for HESS's scope. However, several conceptual and presentation issues should be addressed before the manuscript can be considered for publication. My review focuses primarily on the hydrogeological conceptual framework and broader scientific framing. I leave the detailed technical assessment of the geodetic processing to reviewers with specific expertise in those methods. My comments are provided below.
- The manuscript is currently framed predominantly as a case study of the Sierra Nevada-Central Valley system. However, the scientific problem addressed here is considerably broader and is relevant to mountainous hydrogeological systems worldwide. From my reading, I interpreted the following research questions: "Is Sierra Nevada snowmelt hydraulically connected to aquifers in California's Central Valley through mountain-block recharge, and can this connection be detected from the relative timing of hydrological and geodetic signals?" I encourage the authors to reformulate this question in terms of the broader process rather than the specific study area. For example: "Can a snowmelt-driven hydraulic signal generated in mountain systems be detected in adjacent sedimentary aquifers using integrated hydrological and geodetic observations?" And then the authors can go even more in-depth by asking (or something similar): if so, how can different remote sensing techniques capture this signal? Such framing would substantially increase the broader relevance of the manuscript and would allow the Sierra Nevada-Central Valley system to be presented as a natural laboratory in which this general hypothesis is tested.
- The broader framing proposed above should be supported by a more geographically balanced review of the mountain hydrogeology literature. The current introduction focuses predominantly on studies from the western United States, whereas hydraulic connections between mountain blocks and adjacent sedimentary aquifers have been investigated in a variety of geological and climatic settings worldwide. The conceptual basis of regional groundwater circulation can be traced to classical groundwater-flow theory (e.g., Tóth, 1963), while mountain-front and mountain-block recharge have subsequently been developed as specific conceptual frameworks for mountain-to-aquifer transfer (e.g., Wilson and Guan, 2004; Markovich et al., 2019). This interflow has since been investigated in several ranges worldwide (e.g., Peng et al., 2018; Taucare et al., 2020; Fu et al., 2022; Armengol et al., 2024; Bouimouass et al., 2024; Parizot et al., 2025), while other studies have specifically examined the role of snow and the cryosphere in mountain groundwater recharge and storage (e.g., Carroll et al., 2020; Jódar et al., 2021; Lone et al., 2021; Taucare et al., 2025; Valois et al., 2025; Rispal et al., 2026). Placing the present study within this broader literature would help the scientific gap emerge more naturally: if mountains are hydraulically connected to adjacent downstream alluvial aquifers, and snowmelt represents an important input to these hydrogeological systems, can the resulting hydraulic signal be detected downstream using an integrated multi-sensor observational framework? In my view, this broader question better captures the study's potential novelty and general relevance. I left below some references to complete the literature review:
- Armengol S. et al., 2024. Isogeochemical Characterization of Mountain System Recharge Processes in the Sierra Nevada, California.
- Bouimouass H. et al., 2024. The importance of mountain-block recharge in semiarid basins: An insight from the High-Atlas, Morocco.
- Carroll R.W.H. et al., 2020. Baseflow Age Distributions and Depth of Active Groundwater Flow in a Snow-Dominated Mountain Headwater Basin.
- Fu Y. et al., 2022. Characteristics of hydraulic conductivity in mountain block systems and its effects on mountain block recharge.
- Jódar J. et al., 2020. Snowmelt as a determinant factor in the hydrogeological behaviour of high mountain karst aquifers.
- Lone S. et al., 2021. Meltwaters dominate groundwater recharge in cold arid desert of Upper Indus River Basin (UIRB), western Himalayas.
- Markovich K.H. et al., 2019. Mountain‐Block Recharge: A Review of Current Understanding.
- Parizot O. et al., 2025. 3D Structural Modelling and Restoration of a Deformed Alpine Karst Reservoir.
- Peng T.R. et al., 2018. Assessing the recharge process and importance of montane water to adjacent tectonic valley-plain groundwater.
- Taucare M. et al., 2020. Groundwater resources and recharge processes in the Western Andean Front of Central Chile.
- Taucare M. et al., 2025. High-mountain groundwater quality affected by natural acid drainage.
- Toht J. 1963. A theoretical analysis of groundwater flow in small drainage basins.
- Valois R. et al., 2025. Role of Sparsely Glacierised Basins and Groundwater in Semiarid Andes Periglacial Headwaters.
- Rispal N. et al., 2026. Impact of seasonal snow on the recharge of a mountain karst aquifer under climate change.
- Wilson J. and Guan H. 2004. Mountain-Block Hydrology and Mountain-Front Recharge.

- Following the previous comments, I also suggest reconsidering the title. In addition to its strongly location-specific framing, the current title places considerable emphasis on the "deep aquifer". However, after reading the manuscript, I do not interpret aquifer depth as the primary scientific subject of the study. Rather, the deeper and locally (semi-) confined aquifer system appears to provide a particularly useful natural laboratory because it is expected to be less directly affected by local and shallow recharge processes, thereby allowing the authors to search for a hydraulic signal potentially transmitted from the mountain system. This distinction is subtle but important: the scientific target is the propagation and detection of the mountain-derived hydrological signal, whereas aquifer depth is primarily a condition that facilitates its observation. Moreover, defining "deep aquifers" using a threshold of >50 m is inherently system-dependent and may unnecessarily limit the study's general applicability. A more process-oriented title could therefore be considered, for example: "Tracing Snowmelt-Driven Hydraulic Signals to Assess Mountain-to-Basin Aquifer Connectivity Using Multi-Sensor Observations".
- The Introduction should be substantially reorganised around the scientific problem rather than around the specific characteristics and water-management issues of the study area. In its current form, the section is overly long and structurally diffuse, moving between groundwater overexploitation, managed aquifer recharge, the hydrogeology of the Central Valley, mountain-block and mountain-front recharge, climate change, and remote-sensing techniques without a sufficiently clear progression towards the research gap. As a consequence, the main scientific question becomes apparent only relatively late in the section. I strongly encourage the authors to shorten and restructure the Introduction following a deductive, problem-oriented narrative. A possible organisation would comprise four focused paragraphs: (i) introduce the broader hydrogeological problem, i.e., the challenges associated with understanding recharge, groundwater circulation, and mountain-to-aquifer connectivity in complex mountainous systems; (ii) develop the state of the art on the role of snowmelt in mountainous hydrogeological systems and explain why tracing this input into adjacent aquifers remains difficult; (iii) introduce the methodological gap and explain what an integrated multi-sensor approach combining hydrological and geodetic observations can reveal that individual techniques cannot, particularly regarding the detection and propagation of hydraulic signals; and (iv) present the research question and objectives, followed by a concise description of the study area and methodological approach, and clearly state the expected contribution and novelty of the study. Case-specific information that is not necessary to establish the scientific problem, particularly detailed discussion of groundwater exploitation, managed aquifer recharge, and Central Valley water management, could be substantially reduced or transferred to a new Study Area section where appropriate.
- The Methods section would benefit substantially from the inclusion of a workflow diagram or conceptual methodological scheme. In its current form, the section is rather dense and, at times, overly detailed regarding standard processing steps, making the study's overall analytical logic more difficult to follow. A schematic figure summarising the main datasets, the preprocessing applied to each, the extraction of annual signals and peak timing, the comparison between hydrological and geodetic observations, and the final first-order pressure-diffusion assessment would greatly improve readability. Such a figure would help readers quickly understand how the different methodological components are linked and would likely allow the text itself to be shortened and streamlined.
- The Results section contains several interesting observations, particularly the relative timing of mountain hydrological inputs, groundwater levels, deformation, and GRACE-derived groundwater storage. However, I suggest reorganising the section so that these results are presented more explicitly around the main research question rather than around individual analytical products. At present, considerable attention is given to wavelet characteristics and multi-year variability before the central mountain-to-aquifer timing relationship is introduced. In my view, the strongest result of the study is the apparent lag between the mountain hydrological signal and the groundwater response in the downstream aquifer, followed by the relationship between groundwater levels, deformation, and storage. I therefore encourage the authors to place these observations at the centre of the Results narrative and move secondary spectral interpretations or climate-related explanations to the Supplementary Material or Discussion where appropriate.
- More importantly, the groundwater level results should be disaggregated according to hydrogeologically meaningful subsets. The >50 m depth criterion encompasses wells with substantially different confinement conditions, while it seems that only a limited number of observation wells in the southern Central Valley are located below the Corcoran Clay. Because the interpretation of the observed hydraulic signal relies strongly on the assumption that the selected groundwater records are relatively isolated from shallow and local recharge processes, I consider it important to demonstrate whether the inferred timing relationships persist when analysing the most clearly confined wells separately (e.g., below-Corcoran wells), as well as across major hydrogeological subregions. Such an analysis would provide a much stronger test of the central hypothesis than a basin-wide aggregated groundwater level distribution.
- The current separation between Results and Discussion generates considerable repetition, particularly because several observations described in Section 3 are restated in Sections 4.1-4.4 before being interpreted. In this manuscript, I believe that a combined Results and Discussion section would provide a clearer and more efficient narrative. The study is fundamentally built around a sequence of linked observations and the identification of the mountain hydrological signal, its apparent detection in the downstream alluvial aquifer, the relative timing of groundwater levels, soil deformation, and storage, and, finally, the physical plausibility of pressure transmission. Presenting each observation together with its interpretation and associated uncertainties would make the argument considerably easier to follow and would reduce redundancy. I therefore suggest reorganising the section around the main scientific questions rather than around individual datasets or analytical steps.
- The current "Summary & Outlook" section is also unnecessarily long and partly reiterates material already discussed in Sections 4.1-4.5. I suggest replacing it with a concise Conclusions section that highlights the study's key takeaways. A short set of clearly stated conclusions, potentially presented as bullet points, could summarise: (i) the detected lag between the mountain hydrological input and the groundwater-level response; (ii) the correspondence between groundwater levels and deformation; (iii) the later GRACE-derived groundwater-storage response; (iv) the physical plausibility indicated by the first-order diffusion analysis; and (v) the extent to which these observations support mountain-to-aquifer hydraulic connectivity and are consistent with, but do not uniquely demonstrate or quantify, mountain-block recharge. The broader methodological contribution of the multi-sensor framework could then be stated in a final sentence.

Citation: https://doi.org/10.5194/egusphere-2026-3061-RC3

Susanna Werth, Manoochehr Shirzaei, Grace Carlson, and Roland Bürgmann

Supplement

https://doi.org/10.5194/egusphere-2026-3061-supplement

Susanna Werth, Manoochehr Shirzaei, Grace Carlson, and Roland Bürgmann

Viewed

Total article views: 199 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
138	50	11	199	28	5	6

HTML: 138
PDF: 50
XML: 11
Total: 199
Supplement: 28
BibTeX: 5
EndNote: 6

Views and downloads (calculated since 11 Jun 2026)

Month	HTML	PDF	XML	Total
Jun 2026	60	16	5	81
Jul 2026	47	26	6	79
Aug 2026	31	8	0	39

Cumulative views and downloads (calculated since 11 Jun 2026)

Month	HTML	PDF	XML	Total
Jun 2026	60	16	5	81
Jul 2026	47	26	6	79
Aug 2026	31	8	0	39

Viewed (geographical distribution)

Total article views: 149 (including HTML, PDF, and XML) Thereof 149 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 18 Aug 2026

Short summary

As droughts intensify, groundwater recharge is critical for sustaining agriculture in California’s Central Valley. Using geodetic remote sensing, groundwater, and snowmelt observations, we identify seasonal links between Sierra Nevada snowpack and deep aquifer responses in the valley. The results suggest that mountain groundwater systems contribute to deep valley recharge and should be included in models used for drought planning and groundwater management.


Total:	0
HTML:	0
PDF:	0
XML:	0