Comments for essd-2026-18

This study presents a catalog of extreme seasonal rainfall events in northern Chile during 1979–2019. Using observation and reanalysis datasets, the authors classify extreme precipitation events across three latitudinal bands. The topic is relevant and potentially useful for understanding extreme precipitation and related hazards in the arid and semi-arid regions of northern Chile. The construction of an event catalog could provide a valuable resource for future storyline analysis, precipitation-pattern identification, and hazard assessment. However, the manuscript still requires substantial clarification and improvement before publication. My specific comments are as follows:

Lines 25–28: This paragraph would be better placed around Lines 44–51. Please combine these two paragraphs to improve the logical flow of the introduction.

Lines 91–93: Please explain the rationale for selecting the three latitudinal bands: 17.5°S–23.5°S, 23.5°S–27°S, and 27°S–30°S. Are these bands based on climatic regimes, topography, hydrological basins, previous literature, or the spatial distribution of extreme events?

Line 98: Please clarify the unit used in Figure 1b–d. In the text, the threshold is described as 10 mm/day, whereas the figure appears to use mm as the unit. This inconsistency should be resolved.

Figure 1: Please label the panels as a, b, c, d instead of using the current leftmost and subsequent panel descriptions. In addition, the heights of the colorbars are inconsistent and should be standardized.

Lines 109–110: I am concerned about the definition of extreme rainfall events. If the maximum rainfall day is identified independently at each grid cell, the maximum day at one grid cell may differ from that at a nearby grid cell. Therefore, the identified “event” may not correspond to a single actual weather system. Under this definition, it is difficult to analyse the spatial and temporal extent of extreme precipitation events. The authors should clarify whether the catalog is based on grid-cell maxima, latitudinal-band maxima, or coherent weather systems.

Figure 1 / Figures 3 and 4: In Figure 1, the authors state that LB3 shows the strongest precipitation events in winter, JJA. However, Figures 3 and 4c–d focus on LB2 during the winter season. Please explain why LB2 was selected for the winter-season analysis instead of LB3.

Table 1: The sequence of precipitation categories is not consistent among different latitudinal bands. Please use a consistent category order across all latitudinal bands to make the table easier to compare.

Figure 5: I suggest redesigning this figure to make the comparison clearer. The two rows could be combined, and each bar could be divided into components representing the subcategories shown in the second row. For example, for LB1 DJF, the bar could be separated into North Andes, Along Andes, and South Andes components. 

Line 279: Please justify the use of 4000 m³ as the threshold for debris-flow volume. 

General Comments:

This manuscript presents a catalog of seasonal extreme precipitation events in northern Chile (17.5°S–30.0°S) from 1979 to 2019. By integrating data from CR2METv2.5, ERA5 IVT, MODIS snow cover, and a landslide/debris flow inventory, the study conducts a systematic analysis of event types, snowline elevation, and applications in disaster management. The dataset is regionally valuable and potentially useful for studies of extreme precipitation, moisture transport, snowline variability, and hydro-geomorphic hazards in arid and semi-arid mountain environments.

However, the manuscript does not yet provide sufficient detail regarding event selection, classification criteria, dataset structure, quality control, and uncertainty assessment. Some key classifications rely on subjective interpretation, and the inferences regarding applications to debris flows appear somewhat over-stretched. We recommend a major revision before considering acceptance.

Specific Comments:

1. The manuscript states that they selected the maximum daily precipitation event for each season, but they did not fully explain the specific definition of “extreme events”. Does this mean that there is one extreme event per grid cell each year? Why was it defined this way? The description is very vague. Furthermore, this criterion ignores event duration, accumulated precipitation, and spatial extent, all of which are critical for hazard triggering.

1. Due to the sparse network of stations and complex topography in the mountainous regions of northern Chile, this catalog—which is primarily based on daily-resolution grid data from CR2METv2.5 and ERA5—may underestimate localized short-duration heavy precipitation. ERA5 also has resolution limitations in its IVT representation near complex terrain. It is recommended that the authors conduct cross-validation using station observations, historical events, or other precipitation products.
2. The categories Andes Mountain, Coast, Stratiform, and their subcategories are central to the paper, but their definitions remain qualitative. The authors should provide quantitative criteria based on precipitation location, precipitation fraction by geomorphic zone, precipitation centroid, IVT direction, IVT convergence, or moisture-source contribution. If expert judgement was used, the workflow and classification uncertainty should be documented.
3. The use of both empirical and MODIS-based methods for estimating the snowline is a key component of this dataset. However, in certain seasons and latitudinal zones, there are significant discrepancies between these two methods, with differences sometimes reaching hundreds of meters or even exceeding 1,000 meters. This has important implications for runoff generation and the interpretation of debris flows. Therefore, the authors should quantify method-related biases, root mean square error (RMSE), seasonal dependence, and classification dependence.
4. The Data Availability section provides only the Zenodo DOI, which is insufficient to describe the data product. We recommend adding a data description table that lists the filenames, formats, variables, units, spatial resolution, time range, missing value coding, coordinate system, event IDs, the meanings of classification labels, and the relationships between files. It would be best to include a simple example of how to read the data, illustrating how to filter by a specific latitude band, season, and event type.
5. The conclusion currently focuses primarily on restating results and describing applications. It is recommended to clearly summarize the total number of events in the dataset, key variables, primary added value, applicable scenarios, scenarios where the dataset is not applicable, and future directions for expansion.

Line-by-line comments

1. The term “Comprehensive Catalog” in the title is strong. The abstract should state the total number of events, included variables, and data-product structure to justify this wording.
2. The abstract states that composite analysis was performed for austral summer events in LB3, but it does not clearly state the main finding or scientific implication. A concise result statement should be added.
3. The abstract mentions comparison with a landslide/debris-flow inventory, but does not report the key finding. The authors should state how many events were matched or what this comparison demonstrates.
4. The abstract uses “Coastal,” while the main text uses “Coast.” Please use consistent terminology throughout the entire manuscript.
5. The introduction provides a fairly comprehensive overview of extreme precipitation and debris flow risks, but does not specifically address the shortcomings of existing event catalogs or databases.
6. The three latitudinal bands are said to be based on precipitation climatology, event frequency, and intensity, but the specific boundaries at 23.5°S and 27°S require clearer justification.
7. Figure 1 shows topography and seasonal precipitation composites, but the sample size for each composite is not indicated. Please add event numbers in the caption or text.
8. Explain why the CR2METv2.5 precipitation product was chosen over station observations, the ERA5 precipitation product, or satellite products. The data section should discuss the product’s advantages, applicability, and limitations.
9. CR2MET may be weakly constrained by stations in arid and mountainous northern Chile. The authors should discuss its reliability for extreme precipitation in the study area or explicitly acknowledge this uncertainty.
10. Line 104-108: ERA5 IVT is analyzed at 20 UTC to capture afternoon convective activity, but not all extreme events necessarily peak at that time. The authors should explain whether other time steps or daily mean IVT were tested.
11. Line 109: The definition of “maximum seasonal precipitation event” remains unclear. Please clarify whether one event is selected per band-season-year based on grid-cell maximum, regional maximum, or area-mean precipitation.
12. Line 115: The manuscript states that negligible or atypical events were excluded, but no thresholds or numbers are provided. Please report the exclusion criteria and the number of excluded events by band and season.
13. The criteria for Andes Mountain, Coast, and Stratiform events should be quantified, for example using precipitation centroid, elevation-band distribution, coastal/mountain precipitation fraction, or spatial coverage.
14. Figures 3 and 4 show classification examples, but it is unclear whether these are typical, random, or borderline cases. Please clarify the selection rationale.
15. Several instances of “W esterly IV T” contain unintended spaces. Please correct them to “Westerly IVT”.
16. Figure 6 shows differences between the two snowline methods, but the interpretation remains descriptive. A boxplot or scatterplot of empirical minus MODIS snowline estimates would be useful.
17. MODIS covers only 2000–2019, whereas the catalog spans 1979–2019. The reliability and limitations of snowline estimates for 1979–1999 should be stated more clearly.
18. Section 5.1 presents composite maps of precipitation, geopotential height, and IVT for the LB3 summer Convergence category, but no significance testing (e.g., t-test or Monte Carlo simulation) is provided.
19. Section 5.2 compares the catalog with the SERNAGEOMIN inventory, but the matching rule is insufficiently described. Please specify whether matching is based on the same day, a multi-day window, or storm episodes.
20. The statement that six storms account for 37% of debris-flow events should be framed cautiously. It indicates that the catalog captures some high-impact storms, not that it represents all hazard-triggering rainfall.
21. The conclusions should more directly summarize the data contribution, including total event number, variable list, category counts, and reuse scenarios.
22. The conclusions should explicitly state key limitations, including daily rainfall resolution, classification uncertainty, snowline-estimation errors, and incomplete hazard inventories.