General comments:

This study presents the WEST volume model, an updated postfire debris-flow volume model that uses an expanded dataset, including events from northern California and Washington. The model also introduces a rainfall ratio variable that normalizes rainfall intensity by the 1-year recurrence interval, improving model performance and transferability across diverse geographic regions. The model evaluation is extensive, assessing performance against the full volume database, regional subsets, prior models, and data-limited regions across the western United States, with performance metrics clearly presented and effectively illustrated. I particularly liked how the authors considered the prior sediment volume models and made a careful effort to build upon the previous work. The improvement in predicting sediment volumes is a significant contribution for postfire hazard assessments. Below I have listed comments that I think can be easily addressed.

Specific comments:

1. Regarding the final model selection, I noticed that models #1 and #3 in Table S3 include the same predictor variables with identical performance metrics, differing only in the use of the i30 vs. i60 rainfall ratio. Table 2 also indicates that both rainfall ratios have the same correlation coefficient (-0.13). Consider clarifying whether the selection of the i30 model was primarily based upon its sub-hourly duration. Additionally, how should practitioners proceed in situations where i30 data are unavailable, or where i60 may be more appropriate for operational forecasting applications?
2. Consider including Gatwood et al. (2000) in your statement on line 79, “There are several postfire debris-flow volume models that include rainfall variables,” especially as Gatwood et al. (2000) uses the maximum 1-hour precipitation.
3. I think that the statement surrounding the standard deviation on line 499 is not in agreement with the standard deviation values listed in Table 6. The in-text statement reads “The WEST model slightly overpredicted volumes from data-limited regions but had the lowest standard deviation of the four models (Table 6).” I believe Table 6 lists the lowest standard deviation from the data-limited regions as 1.29 (EAV model) not the WEST model (1.37).
4. “Rainfall ratio” and “rainfall anomaly” are used interchangeably throughout the manuscript and the supplemental. For example, on line 415, “rr” is used in Eq. 2, whereas “ra” is used in Table S3. Although everything is clearly defined, maintaining consistent naming of this variable throughout the manuscript and supplemental may improve clarity for the reader.
5. Although Table 1 displays the number of volume measurements geographically, it lacks information surrounding the magnitude (e.g., 10¹ – 10⁴) of the sediment volumes geographically. Consider adding a column with the range of sediment volumes in Table 1.
6. While the need for additional volume data from data-limited regions is discussed, it may be helpful to also acknowledge the importance of including larger-magnitude sediment volumes, as the current dataset from these regions contains smaller volumes.

Technical Corrections:

• Table 1 – consider adding more text for context to the Table 1 caption (e.g., “Fire information associated with measured sediment volumes”)
• Line 123, add space after “,” in “Smith et al.,2021”
• Figure 1 – consider simplifying the fire icon to a circle symbol for improved legibility and/or lighten the ecoregion colors to increase contrast between the ecoregion and the volume symbol.
• Line 256 – for increased clarity, consider including the text “for each watershed” in the phrase “We calculated another 17 variables for each watershed related to fire characteristics (Table 4)…”
• Line 271 – delete “and” directly after “B_k”
• Line 415 – consider spelling out the first instance of “VIF” in this section for clarity.
• Lines 457, 462, and 523 – change “Figure S2” to “Figure S3”
• Line 460 – delete “a” directly after “MAE (Table 7) and”
• Line 489 – change “Figure S5” to “Figure S5a” to maintain consistency with “Figure S5d”
• Line 657 – change “Figure 6” to “Figure 7”

References:

Gatwood, E., Pedersen, J., & Casey, K. (2000). Los Angeles district method for prediction of debris yield. US Army Corps of Engineers, Los Angeles District, 145.

Thanks for the opportunity to review this MS. The MS describes the development of an empirical model of post fire debris flow volume that draws on a greater quantity of observation data and aims to create a more general model for the Western US than the current models developed from datasets collected in different regions. The paper is well suited to the audience of NHESS. The overall objective is sound ( a more generic model), The methods are well suited to the aim and are executed in an extremely structured and well defended and explained way. I particularly liked the extensive detail justifying the logic of the decisions regarding the acceptance/exclusion of the many possible models. The performance comparisons with existing models were really well justified in terms of the metrics used but also the graphics and tables. The authors conclusions and interpretation of the results were clear and concise. Overall this is a very high quality MS. I only have two minor suggestions. One relates to the figures of distributions of the residuals; would it be possible to include a second x axis with the untransformed values for this distribution, as it is difficult to interpret in the context of units that are intuitive. The second point is, i know the scope of the model development is restricted to the western US, but it would be good to see the results interpreted a little in the context of the rest of the world. How does this model/modelling approach/variables compare to what others have published on this topic?  What might the implications be for this new anaysis for other researchers with similar objectives in other regions/continents re estimating post fire debris flow volumes? This could make the US work more relevant to a wider audience than western US practitioners and researchers.