| 01 Oct 2026
Dynamic critical groundwater depth as a predictor of irrigation-intensified salinization in lowland Hungary
Abstract. Shallow groundwater in continental lowland environments sustains upward capillary fluxes that transport dissolved salts to the land surface. However, the depth below which this capillary-driven contribution becomes negligible, often parameterized as the extinction depth in groundwater model ET packages, has been treated as a static, soil-dependent parameter. We argue that salinization risk is governed by the dynamic overlap between the seasonal cycles of groundwater level (GWL) and evapotranspiration (ET), and develop a time-varying risk index based on the proximity of the water table to a dynamic, GWL-dependent critical depth, and capillary fringe. The framework is applied to a 4041 km² sub-region of the TIKEVIR system in southeastern Hungary using 74 years (1951–2025) of gridded GWL observations and FORESEE-HUN climate forcing. The domain-averaged seasonal cycle (shallowest in April at 2.50 m, deepest in October at 3.04 m, amplitude 0.54 m) defines a critical salinization window in March–June when the water table resides within the critical zone (zGWL < dcrit) under rising evaporative demand. A ConvLSTM network with eight input channels, trained on 1971–2019 data and evaluated on 2020–2024, achieves NSE = 0.839 and RMSE = 0.416 m for 1-month-ahead GWL forecasting at 1 km resolution, with 27–40 % RMSE reduction over persistence at 1–12-month lead times. Monte Carlo dropout analysis has verified a narrow prediction uncertainty, with a mean risk credible interval width of 0.014 (dimensionless). By integrating the forecast with the risk index, it was determined that 19.2 % of the domain, equivalent to 776 km², is persistently at high or critical risk, predominantly located in discharge zones. A notable secular deepening of the water table, measured at 0.051 m per decade (p = 0.032), has slightly diminished the domain-averaged risk; however, the feedback loop between irrigation and salinization may counterbalance this trend. This methodology holds potential for application to similar lowland aquifer systems where the groundwater level (GWL) and evapotranspiration (ET) cycles are phase-offset.
This manuscript integrates groundwater-level reconstruction, dynamic critical groundwater depth parameterization, and ConvLSTM forecasting to develop a framework for assessing and predicting soil salinization risk. Although considerable effort has been devoted to this study, the core empirical formulations lack sufficient physical justification and independent validation, and their relationship with soil salinization processes remains unverified. Furthermore, the novelty of the proposed approach and its advantages over conventional methods have not been convincingly demonstrated. These fundamental issues substantially undermine the scientific rigor and reliability of the main conclusions. Therefore, I recommend rejection of the manuscript in its current form.
Major comments:
1. Insufficient physical justification and demonstrated novelty of the dynamic critical groundwater depth approach
The manuscript presents the dynamic critical groundwater depth as a major methodological innovation, defining it as an empirical function of groundwater depth in Eq. (1). However, the proposed formulation depends solely on groundwater depth, without explicitly considering the effects of soil hydraulic properties and evapotranspiration (ET) on capillary rise. In particular, the Introduction emphasizes the seasonal phase offset between groundwater levels and ET as an important motivation for this study, yet this relationship is not reflected in Eq. (1). Although the authors collected van Genuchten soil hydraulic parameters, these parameters were not used in calculating the critical depth. Furthermore, the functional form and empirical coefficients (0.8, 0.4, and 3) lack clear physical justification or independent calibration. The authors mainly support the proposed formulation by comparing the calculated critical depths with previously reported values. However, agreement with the reported range does not validate the functional relationship or its dynamic behavior.
Moreover, the manuscript does not demonstrate the advantages of the proposed dynamic critical depth over conventional static critical depth approaches. Even with a fixed critical depth, seasonal groundwater-level fluctuations can produce temporal variations in the calculated risk index. Therefore, the reported seasonal risk patterns and the spatial similarity between critical depth and groundwater depth do not provide independent evidence of the superiority of the proposed method. I suggest that the authors directly compare the dynamic and static critical depth approaches under the same conditions and validate their performance against independent observations. Without sufficient physical justification and comparative validation, it remains unclear whether the proposed method represents a meaningful scientific advance or simply a nonlinear transformation of groundwater depth.
2. Insufficient physical justification and validation of the salinization risk index
Equation (5) introduces a salinization risk index based on groundwater level (GWL) depth, dynamic critical depth, and capillary rise height. However, the physical basis for the proposed formulation and its linear interpolation remains unclear. Since both the dynamic critical depth and capillary rise height are defined as functions of GWL depth, the resulting risk index is essentially a nonlinear transformation of GWL depth. Moreover, the index does not explicitly consider groundwater salinity, soil salinity, or actual salt transport and accumulation. Given that areas with similar GWL depths may exhibit different degrees of salinization, the authors need to clarify how this empirical index can reliably represent actual salinization risk.
More importantly, the proposed risk index has not been independently validated against observed soil salinity or salinization patterns. Section 4.3.6 only examines correlations with clay and CaCO₃ contents, which do not directly represent soil salinization. Furthermore, the sensitivity analysis in Section 4.4 shows that the scaling parameter (kd) strongly affects the calculated risk, yet the baseline value of (kd = 2.5) m lacks sufficient calibration. The capillary rise parameter (kc) and the risk classification thresholds (0.1, 0.3, 0.5, and 0.7) also lack clear justification. Therefore, the reported 19.2% of the study area classified as high or critical risk, as well as the spatial and seasonal risk patterns, depend strongly on the assumed formulations and parameter choices. Without independent validation and reliable parameter calibration, these results cannot be considered robust evidence of actual soil salinization risk.
3. Groundwater-level forecasting accuracy does not establish the reliability of future salinization risk predictions
This manuscript uses ConvLSTM to predict future groundwater levels (GWL) and then calculates salinization risk using the empirical risk index. However, the relationship between GWL depth and soil salinization has not been adequately validated. Therefore, accurate GWL predictions do not necessarily imply reliable salinization risk forecasts. Although ConvLSTM shows good performance against the reconstructed GWL grids, the point-scale validation against monitoring wells in Section 4.5.5 shows considerably poorer results, raising concerns about the reliability of the predictions. Furthermore, the projections to 2050 extend far beyond the evaluated forecast horizons of 1-12 months. Without independent validation of the salinization risk index and sufficient evidence supporting long-term GWL forecasts, the projected future salinization risk remains highly uncertain.
4. Lack of quantitative evidence for the proposed irrigation-salinization feedback
The manuscript title and Discussion emphasize that irrigation may intensify soil salinization by increasing groundwater recharge, raising groundwater levels, and enhancing capillary rise. However, irrigation amount and timing are not explicitly included in the ConvLSTM inputs, and the proposed framework does not simulate irrigation-induced changes in groundwater levels or salt transport and accumulation. Therefore, the irrigation–salinization feedback has not been quantitatively demonstrated through irrigation scenarios or independent observations. The discussion of this feedback remains largely conceptual and does not provide sufficient evidence to support the emphasis placed on irrigation-intensified salinization in the title and main conclusions.