The authors aim to explore the definition and detection methods of drought-to-flood transitions (DFTs), and emphasizes the limitations of existing threshold-based approaches in detecting such consecutive extreme hydrological events. Using eight case study catchments, authors compare the detection effectiveness of three threshold methods (fixed, seasonal, and dynamic thresholds) for DFT events and verifies the results by referring to media reported disaster events. The topic is interesting and meaningful, but it also has certain limitations. My comments are outlined below.

Major comments

1. Although the title of this manuscript is mentioning "what is a drought- to-flood transition", in the definition of drought-to-flood transitions(2.3.2), I only find the definition of the transition time. There are no restrictions on the duration of droughts and floods. Long - term droughts, short - term droughts, or interrupted droughts have different effects and need clear definitions. Therefore, I suggest that the authors include additional content to address and clarify this point.
2. Figure 6 shows that different threshold methods vary significantly in seasonal catchments, but the mechanism is unclear. I suggest adding how dynamic thresholds suppress or amplify seasonal anomalies.

Minor comments

1. Some cases ( Chilean and Italy) have short data periods (under 20 years), which may affect the stability of threshold calculations. This needs to be clearly stated in the limitations.
2. The "90 days window" for DFT is based on prior studies, yet its applicability across different climate zones (tropical vs. temperate) isn't discussed. It's recommended to add a sensitivity analysis.
3. When identifying DFT events, three different drought threshold approaches were used. The paper mentions "calibration" to ensure these methods detect the same number of drought/flood events, but the specific calibration process isn't explained. It's recommended to add technical details or cite relevant literature.
4. Fig 2. Figure 2b has no label “b”.

Review comments for “What is a drought-to-flood transition? Pitfalls and recommendations for defining consecutive hydrological extreme events” by Anderson et al. 

The present study provides a timely and well-motivated investigation into the definition and identification of drought-to-flood transition event, which is an increasingly important topic in hydrology. By comparing multiple threshold-based methods across eight case study catchments, the authors demonstrate how methodological choices can significantly influence event detection and interpretation. These findings carry potential implications for understanding compound and consecutive hydrological hazards. However, several methods appear to introduce potential biases. In addition, the study lacks a strong physical grounding, and certain sections would benefit from clearer justification and interpretation to enhance scientific rigor and applicability. Please see my detailed comments below:

1. Abstract: The current abstract is primarily qualitative in nature and may benefit from the inclusion of key quantitative findings to strengthen its scientific clarity.

2. L106-108: The reliance on streamflow thresholds without incorporating standardized indices may limit comparability with other studies and hinder the evaluation of meteorological or soil moisture-driven processes. The authors should more clearly justify this choice and discuss its implications for generalizability.

3. L140-143: The approach of defining drought and flood periods based on “reported” start and end dates from media or scientific sources may introduce a degree of subjectivity, particularly as media reports can vary widely in temporal resolution and geographic specificity. Did the authors use any verification measures to address potential reporting bias or spatial/temporal mismatches? It is also recommended that these reported dates be supplemented with objective hydroclimatic metrics to enhance consistency.

4. L161-162: This limited data coverage (17–20 years) may compromise statistical robustness of this study. Did the authors perform any sensitivity checks to assess the influence of record length on event detection and methodology performance?

5. L169-174: Here, the rationale for choosing the specific percentile window (±15 days) and the smoothing approach (31-day rolling mean) may need further justification.

6. L194-196: Calibrating threshold percentiles to match target event frequencies may obscure the physical hydrological meaning of the thresholds. These calibrated percentiles may not align with widely accepted definitions of drought and flood conditions, potentially limiting comparability with other studies. The reviewer suggests that the authors further discuss its potential biases and implications for physical interpretation.

7. L207-209: The fixed 90-day interval for defining transitions oversimplifies the wide range of hydrological responses across diverse catchments and climatic regimes. The authors should justify the applicability of this threshold across all case studies and conduct sensitivity analyses to assess how the chosen interval affects event detection.

8. Line 230: The choice of the 5th percentile as a transition threshold lacks clear justification. Besides, while bootstrapping accounts for sparse or uneven data, it may also smooth over physically meaningful variability between catchments. The reviewer recommends discussing the sensitivity of results to this resampling approach and analyzing how catchment-specific hydrological characteristics influence the derived thresholds.

9. L235-236: The reviewer suggests the authors clarify the criteria used to distinguish between transitions "not of interest" and those "potentially representing negative impacts." This distinction appears subjective and would benefit from being more explicitly linked to hydrological or societal impact indicators.

10. L391-394: While the authors stress methodological influence, it could more clearly relate detection differences back to physical processes such as antecedent soil moisture and rainfall intensity to aid in bridging the gap between statistical detection and hydrological realism.