The manuscript presents a assessment of low flow freq analysis using non-stationary extension of Weibull and copula based dependence framework.  The framework is used for evaluating around 700 European catchments from the ROBIN dataset and demonstrates the implications of non-stationarity on low-flow frequency estimates, particularly for rare events. The study addresses an important topic given increasing concerns regarding climate-change induced nonstationary impacts on hydrological droughts and low-flow conditions. The manuscript is generally well structured and methodologically rigorous. However, some aspects require  a bit of attention and explanation before publication. Here are some points to consider by the authors:

1. The introduction provides a good overview of non-stationary frequency analysis literature, however authors should more clearly explain why existing approaches (e.g. GAMLSS-based low-flow frequency analysis) are insufficient and specifically why the proposed extension of LFAmix and LFAmixC is necessary in the set up which they have used.
2. The manuscript adopts WEI3 throughout, but should the justification or advantages/disadvantages of using alternative distributions be included. Since low-flow frequency analysis often employs Lognormal, Generalized Logistic, Pearson Type III or GEV formulations, authors should discuss why WEI3 is uniquely suitable and whether results are sensitive to this choice.
3. Non-stationarity is only introduced through the location parameter. While the authors justify this using parsimony arguments, climate change can influence both variability and shape of low-flow distributions. Understandable that this will make the model very more sensitive and risky….some discussion or such probabilistic sensitivity analysis is recommended.
4. The framework assumes a linear temporal evolution in location and dependence parameters. However, low-flow behaviour often exhibits regime shifts, accelerations, thresholds, or oscillatory behaviour linked to climatic variability. If not nonlinear relationships related analysis is done, at least this should be discussed in detail.
5. The use of PAM clustering and seasonality histograms is interesting, but the physical interpretation of the resulting clusters needs stronger discussion. For example, how sensitive are the results to the chosen clustering approach, silhouette threshold (0.2), or seasonal definitions (April–October and November–March)?
6. WEI3 fitting was successful in only 76.1% of catchments. This implies nearly one-quarter of stations could not be fitted successfully. May be the characteristics of these excluded catchments should be analysed and discussed, as this may introduce regional bias into the results.
7. The paper focuses heavily on one Austrian catchment. Authors should consider presenting at least one additional example, preferably from a different characteristics dominated regime, to demonstrate framework performance across contrasting hydro-climatic conditions.
8. Statements such as a 100-year event becoming a 43-year event are powerful but may be misinterpreted by readers. It should be clarified that these changes are statistical diagnostics based on the fitted non-stationary model and should not be interpreted directly as future projections.