| 12 May 2026
Influence of synoptic patterns (NAO vs. WeMO) on rainfall isotopic composition in SE Iberia: A machine learning approach
Abstract. Atmospheric circulation controls moisture transport across the western Mediterranean, but its seasonally resolved imprint on precipitation isotopes in many areas remains poorly understood, limiting the integration of circulation diagnostics with modern isotope monitoring. Southeastern Iberia constitutes an outstanding natural laboratory to identify and quantify the roles and isotopic footprint of the main modes of atmospheric variability in the region (i.e., the Western Mediterranean Oscillation (WeMO) and the North Atlantic Oscillation (NAO)) and other local factors, such as altitude and precipitation amount. Here, we combine a multi-altitude (560–1800 m a.s.l.) precipitation-isotope network in Sierra de Segura, a mountain range located in SE Spain with explainable machine-learning methods to quantify how large-scale circulation and precipitation regime control rainfall isotopic composition.
First we extend the WeMO index forward from 2020, when instrumental measurements end, to 2025 using a physically constrained XGBoost reconstruction based on regional sea-level pressure predictors from San Fernando (Spain) and multiple northern Italian stations, reproducing the published index over 2010–2020 with robust cross-validated performance (R² = 0.85 ± 0.05; RMSE = 0.41 ± 0.07, n = 129 months). Then we analyze 448 rainwater samples collected between 2017 and 2023, aggregated into 154 precipitation-weighted monthly observations of δ¹⁸O, δ²H and δ¹⁷O, together with d-excess and ¹⁷O-excess. Bulk isotope ratios co-vary strongly (ρ > 0.95), allowing δ¹⁸O to represent the dominant network-scale signal.
Random Forest models interpreted using SHAP reveal a clear seasonal reorganization of controls. During the wet season (October–March), δ¹⁸O variability is primarily circulation-driven, with the North Atlantic Oscillation acting as the dominant control and the WeMO providing a consistent secondary modulation. In contrast, during the dry season (April–September), δ¹⁸O is governed by precipitation regime, with precipitation amount overwhelmingly predominating over circulation indices and exhibiting a strongly non-linear, amount-effect response. Back-trajectory clustering of regionally coherent isotope-sampling events supports these contrasting seasonal influences, indicating a small number of recurrent synoptic transport regimes in winter and weaker synoptic organization in summer.
Overall, these results provide a transferable, seasonally calibration that links synoptic circulation and precipitation regime to rainfall δ¹⁸O (and excess parameters) in southeastern Iberia. This innovative framework enables robust interpretation of Mediterranean isotope-based paleoclimate archives by distinguishing circulation-driven wet-season signals from precipitation-regime-driven dry-season variability. It is readily applicable to other climatically complex areas influenced by multiple modes of atmospheric variability and pronounced seasonal contrasts. Furthermore, our results provide a basis for assessing the sensitivity of cave and lake systems to future changes in circulation persistence and rainfall intermittency under ongoing climate change.
General assessment
This manuscript addresses a highly relevant topic, focusing on the controls exerted by large-scale atmospheric circulation patterns on precipitation isotopes in southeastern Iberia. The dataset is valuable, the extension of the WeMO index is particularly interesting and may prove useful for a wide range of future studies in the western Mediterranean region. Also, the application of machine-learning techniques provides an innovative perspective for investigating isotope-climate relationships. The study has clear potential significance for both modern hydroclimatology and isotope-based paleoclimate reconstructions in the western Mediterranean region. However, some aspects require substantial clarification before publication. In particular, the physical interpretation of the observed NAO-δ18O relationship remains insufficiently explained and currently weakens the main conclusions of the paper. For this reason, I recommend minor-to-moderate revision.
Major comments
1. The negative NAO-δ18O relationship requires a stronger physical explanation
The most important issue concerns the interpretation of the wet-season relationship between δ18O and NAO. The manuscript identifies NAO as the dominant predictor of δ18O variability during the wet season, supported by multiple analyses. However, the observed relationship is negative, meaning that more positive NAO conditions are associated with more negative δ18O values. This result appears difficult to reconcile with the classical understanding of Mediterranean hydroclimate. In many Mediterranean regions, negative NAO phases are associated with wetter conditions (e.g., Luppichini et al. 2026 and reference therein), which would generally be expected to produce more depleted isotopic values through the amount effect (e.g., Baldini et al. 2008; Comas-Bru et al. 2016) or other processes (higher rainout, different moisture sources, etc.). However, the authors show that precipitation amount is not significantly correlated with δ18O during the wet season, indicating that the observed pattern cannot be explained by a simple amount effect mechanism. If precipitation amount is not responsible for the observed relationship, then the manuscript should investigate which physical processes underlying NAO variability are controlling δ18O. Since NAO is only an index reflecting atmospheric dynamics, the actual climatic driver remains unidentified. Could atmospheric temperature be the main driver behind this effect? The varying contributions of frontal precipitation (under NAO-) vs. convective precipitation could play a role as well. This issue represents the main limitation of the manuscript and deserves a much deeper discussion before publication.
2. Atmospheric temperature is not considered
The second limitation of the study is the complete absence of atmospheric temperature among the predictors used in the statistical and machine-learning analyses. Temperature is widely recognized as one of the principal controls on precipitation isotopes throughout the Mediterranean region and often represents a dominant predictor in continental areas. Temperature should be included among the tested predictors, or the authors should provide a strong justification for excluding it.
3. The sampling strategy and temporal aggregation require clarification
The manuscript does not clearly explain how the isotopic dataset was constructed. Section 2.2.1 (L174-175) states that sampling was conducted every 1-2 months, whereas Section 2.2.2 (L200) refers to "event-based precipitation samples" aggregated into monthly observations. The relationship between the raw samples, event-scale precipitation, and monthly weighted averages remains unclear.
4. Limitations of the trajectory analysis
The HYSPLIT analysis raises some methodological questions. I am not fully convinced that a single trajectory initialized at 12:00 UTC adequately represents the atmospheric pathways associated with an entire precipitation event. In addition, the selected integration time of 96 h appears relatively short compared to many isotope and moisture-source studies, which often use longer trajectories to capture more distal moisture origins. Furthermore, no moisture uptake analysis was performed, making it difficult to identify the actual regions contributing moisture to precipitation at the study sites. How do the authors expect the inclusion of a moisture-source analysis to affect their interpretation?
5. Physical interpretation of the transport-regime results
Several conclusions derived from the trajectory clustering are difficult to reconcile with the expected physical meaning of NAO. In particular, the manuscript repeatedly suggests that Atlantic transport regimes may be favoured under both NAO+ and NAO− conditions. This appears inconsistent with the classical interpretation of NAO, according to which the Atlantic advection is promoted during NAO- conditions and deserves further explanation.
Minor comments
L58-61. The introduction is focused on the Iberian Peninsula. A broader Mediterranean perspective would be useful, particularly considering the growing number of studies, both from a hydroclimatic and isotopic perspective, from other Mediterranean sectors (e.g., Vicente-Serrano et al. 2025, Zhang et al. 2025, Natali et al. 2025, Columbu et al. 2025).
L75. A very recent work from the Italian Peninsula could be useful (Luppichini et al. 2026).
L100 (and throughout the manuscript). Please replace the term “isotopy”. This is not standard terminology in isotope geochemistry. Terms such as isotopic composition, isotope variability, or precipitation isotopes should be used instead.
L174. The citation appears incorrect. The appropriate reference is Natali et al. (2024). Please also clarify whether the paraffin oil was removed before isotope analysis.
Figure 1. The topographic profile shown in panel B should be indicated on the map in panel A.
L193. Please clarify whether isotope normalization was performed using a two-point (VSMOW–SLAP) or three-point calibration. In general, GISP is used as an independent quality-control standard rather than as a calibration point.
L197. Seven injections may be insufficient for high-precision 17O-excess measurements. A larger number of injections (e.g., ten or more) would generally reduce analytical uncertainty.
L201. Precipitation-weighted means are particularly useful for recharge studies. However, if the objective is to investigate atmospheric controls and circulation patterns, arithmetic means may provide a more representative description of atmospheric variability.
L308. The term “monthly observations” remains ambiguous given the sampling strategy.
Section 3.2.2. The negative correlation between δ18O and NAO during the wet season deserves a more thorough discussion, especially considering that precipitation amount does not appear to explain the observed pattern. Conversely, the positive δ18O-WeMO relationship appears physically more consistent with the observed precipitation patterns.
L356–357. The opposite sign of the NAO-δ18O relationship between high and low altitude stations during summer is unexpected and requires further explanation.
L371–374. The positive relationship between precipitation amount and d-excess/17O-excess during the dry season appears physically plausible in terms of sub-cloud processes and convective systems.
Section 3.2.5. The differences among trajectory results at 500, 1500, and 3000 m need to be better explained.
L473. Please include seminal studies on d-excess (e.g., Froehlich et al. 2008) as well as more recent investigations (e.g., Natali et al. 2022).
L501. Interestingly, similar conclusions have also been reported in other recent Mediterranean isotope studies (e.g., Natali et al. 2023, 2025).
L518–521. The interpretation of Atlantic regimes under positive NAO conditions requires further justification.
L536. Interestingly, a similar reduction of the isotopic altitude effect during summer has been reported in Natali et al. 2022.
L546. Again, the relationship between NAO+ conditions and Atlantic transport regimes requires further explanation.
L566–571. The observed behaviour is consistent with several isotope datasets from the Mediterranean region and beyond (e.g., Natali et al. 2021 and references therein).
L600. The discussion of future climate implications relies on the negative NAO-δ18O relationship, which currently lacks a convincing physical explanation.
L616. This interpretation has been reported not only in Spain but also in other Mediterranean regions (e.g., Columbu et al. 2025 and references therein).
References
Columbu A., Isola I., Zanchetta G., Drysdale R. N., Natali S., Hellstrom J. C., Magny M. and Fallick A. E. (2025) Speleothem evidence of late glacial and Early Holocene Preboreal and Boreal hydro-climate changes in western Mediterranean (Corchia Cave, Italy). Quat. Sci. Rev. 354, 109241.
Froehlich K., Kralik M., Papesch W., Rank D., Scheifinger H. and Stichler W. (2008) Deuterium excess in precipitation of Alpine regions – moisture recycling. Isotopes Environ. Health Stud. 44, 61–70.
Luppichini M., Natali S., Columbu A., Zanchetta G. and Bini M. (2026) Mediterranean precipitation variability is driven by complex atmospheric mechanisms beyond the North Atlantic Oscillation. Commun. Earth Environ. 7, 81.
Natali S., Baneschi I., Doveri M., Giannecchini R., Selmo E. and Zanchetta G. (2021) Meteorological and geographical control on stable isotopic signature of precipitation in a western Mediterranean area (Tuscany, Italy): Disentangling a complex signal. J. Hydrol. 603, 126944.
Natali S., Doveri M., Franceschi L., Giannecchini R., Luppichini M., Menichini M. and Zanchetta G. (2023) Moisture sources and climatic effects controlling precipitation stable isotope composition in a western Mediterranean island (Pianosa, Italy). Atmos. Res. 294, 106987.
Natali S., Nigro M., Baneschi I., Giannecchini R., Doveri M. and Zanchetta G. (2024) On the reliability of tube-dip-in-water precipitation collectors in isotope hydrology: A field experiment for low rainfall amounts. J. Hydrol. 644, 132096.
Natali S., Zanchetta G., Luppichini M., Doveri M., Isola I. and Giannecchini R. (2025) Assessing moisture origin as a potential driver of event-based precipitation isotope variability in a Western Mediterranean catchment (Apuan Alps, Italy). Clim. Dyn. 63, 159.
Vicente-Serrano S. M., Tramblay Y., Reig F., González-Hidalgo J. C., Beguería S., Brunetti M., Kalin K. C., Patalen L., Kržič A., Lionello P., Lima M. M., Trigo R. M., El-Kenawy A. M., Eddenjal A., Türkes M., Koutroulis A., Manara V., Maugeri M., Badi W., Mathbout S., Bertalanič R., Bocheva L., Dabanli I., Dumitrescu A., Dubuisson B., Sahabi-Abed S., Abdulla F., Fayad A., Hodzic S., Ivanov M., Radevski I., Peña-Angulo D., Lorenzo-Lacruz J., Domínguez-Castro F., Gimeno-Sotelo L., García-Herrera R., Franquesa M., Halifa-Marín A., Adell-Michavila M., Noguera I., Barriopedro D., Garrido-Perez J. M., Azorin-Molina C., Andres-Martin M., Gimeno L., Nieto R., Llasat M. C., Markonis Y., Selmi R., Ben Rached S., Radovanović S., Soubeyroux J.-M., Ribes A., Saidi M. E., Bataineh S., El Khalki E. M., Robaa S., Boucetta A., Alsafadi K., Mamassis N., Mohammed S., Fernández-Duque B., Cheval S., Moutia S., Stevkov A., Stevkova S., Luna M. Y. and Potopová V. (2025) High temporal variability not trend dominates Mediterranean precipitation. Nature 639, 658–666.
Zhang J., Genty D., Bourges F., L.L. Michel S., Minster B., Régnier E., Devaux L., Bujan S., Su Z. and Laurila T. K. (2025) Source of rainfall above Mediterranean caves (Chauvet and Orgnac) and long-term trend of cave dripping oxygen isotopes based on 20 years monitoring records: Importance for speleothem-based climate reconstructions. Quat. Sci. Rev. 349, 109145.