the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Jul 2025
Seismicity catalogue of the entire Chilean margin (18° to 56°S) from an automated approach
Abstract. In this study we process 5 years of continuous seismic data, from 2017 to 2021, recorded by networks deployed along the Chilean convergent margin of western South America. We select a set of stations so that coverage is as continuous and consistent in time as possible. Thus, changes in the catalogue are mainly due to the processes responsible for the seismicity itself and not network artifacts. For phase picking, we use the deep-learning algorithm Earthquake Transformer, PyOcto for association, NonLinLoc for location and HypoDD for relocation. To handle the large spatial extent of the study area, we create a partitioning workflow that mitigates problems that arise from associator and location codes working in cartesian coordinates. This allows us to obtain a main seismicity catalogue with over 600.000 double-difference relocated events and a completeness magnitude of ~2.5, with events located using an existing 3D velocity model. While this provides better location accuracy than a 1D or 2D velocity model, it limits the final catalogue to the extent of the 3D velocity model which does not cover our whole study area. Therefore, we create a secondary seismicity catalogue for the region to the south of the 3D model. Our catalogues provide a dense, high-resolution dataset, with large spatial extent, for the interpretation of seismic processes along the western South American Margin.
- Preprint
(2608 KB) - Metadata XML
-
Supplement
(1825 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on essd-2025-254', Thorne Lay, 09 Aug 2025
-
RC2: 'Comment on essd-2025-254', Jannes Münchmeyer, 01 Oct 2025
In their manuscript “Seismicity catalogue of the entire Chilean margin (18 to 56 S) from an automated approach”, Riedel-Hornig et al. present a new earthquake catalog covering the 5 year period from 2017 to 2021. With ~600,000 event, this catalog is an order of magnitude large than the routine CSN catalog, providing a detailed view into the seismicity of the Chilean margin. Overall, this is an excellent manuscript. The work is technically sound and the manuscript is well-written and contains all necessary detail. Below I provide a list of comments that should be addressed, however, all of them concern minor points. I recommend publication once these comments have been addressed.
In addition, I wanted to bring up two suggestions. However, I believe both of them are out-of-scope for the manuscript at hand and might rather be of interest for future work. First, I think it would be of great value to extend the catalog to a longer duration. The years since 2021 have good coverage and even analyzing the more sparse data before 2017 with the same workflow would substantially improve upon existing catalogs. The longer duration would allow to systematically study long-term trends. Second, while the location accuracy in the presented catalog is good, it could be improved substantially by incorporating correlation-based differential times, providing detailed insights into fine-scale structures.
Detailed comments:
- Line 81: I’d recommend using “consistent in time” instead of “constant in time” for the catalog. In contrast, the use of “constant” to refer to magnitude of completeness later in the same line seems appropriate.
- Line 84: While I agree on the aspect of temporal changes, being able to identify spatial changes would require uniform station coverage and noise conditions, which are not present here.
- Figure 1: To make part b more informative, I’d suggest sorting the stations by their latitude and scale it to the map on the left. This way, readers can see if certain regions have varying coverage over time.
- For the processing with HypoDD, I’d recommend to explicitly state that no cross-correlation times were used to avoid potential questions.
- For the magnitude calculation, which attenuation function did you use? Is this the same attenuation function used by the CSN?
- For the classification of events, I’m slightly concerned that the tolerance margins might in places be too small to capture the systematic deviation between your catalog locations and slab2. For example, in Figure 3 panel 9, there is a large swath of seismicity between 200 and 300 km that seems mislabeled. Similarly, other research found notable differences in depth between local catalogs and slab2, for example, Sippl et al. (2018), González-Vidal et al. (2023), and Münchmeyer et al. (2025).
- Line 174: Is this the number of associated picks or the total picks? Could you provide the total number of picks before association as well? The same applies to line 230.
- Line 178-180: The observation that the segments 20S-24S and 30S-34S are seismically more active than the region in between should be discussed in a more nuanced way. These segments have substantially denser station coverage than the 24S-30S segment, so the completeness will be lower. I’d suggest estimating completeness individually for each of the segments, truncating at the highest of the three completeness values and see if the change in seismic activity still is significant. As far as I know, the effect should remain intact, but it would be good to make a more solid argument for it.
- For the same segments, I find the argument of incoming topography not particularly convincing. First, the less active segment in between hosts the subducting Taltal and Copiapó ridges, so it raises the question why these don’t correspond to enhanced activity. Second, it seems unlikely that the incoming ridges change seismicity patterns over more than 400 km long segments.
- Figure 2a: Could you add whiskers for the width of the profiles?
- Figure 2: There is a curious cluster of seismicity at depth north of 20S and around 68W. I checked the catalogs from Sippl et al (2018, 2023) but neither of the two report this cluster. It think it would be worthwhile commenting on this and validating whether this is a real feature or a processing artifact.
- Figure 4, panel 3: There is a substantial swath of seismicity well within the oceanic mantle shown. This seems like a surprising feature to me, which is again absent from the Sippl et al catalogs. Again, I’d recommend to comment on this feature and verify whether it is an artifact and what might be causing it.
- In your assessment of absolute uncertainties you report higher errors in latitude than longitude. This is genuinely surprising to me, especially for out-of-network clusters towards the east. Generally, I’d expect uncertainty ellipses here to be elongated in EW directions (perpendicular to the network) and narrow in NS direction (parallel to the network). Could you explain this behaviour?
- Figure 7: Please specify the time frame, i.e., “Number of events per ?”.
- For the segment 24S-30S, you compare to the González-Vidal et al. (2023) catalog. I’ve recently published a newer catalog for this region in Münchmeyer et al. (2025), using a superset of the data used by González-Vidal et al. (2023) and covering a longer duration. Given this catalog has higher resolution and substantially more events, I’d suggest updating the comparison. However, I admit that my catalog was only available as a preprint at the time of submission of this manuscript and I also don’t want to push my own work too much, so I fully appreciate if you decide not to change this comparison.
- Line 322: The statement that the DSZ is continuous throughout the Chilean margin is an oversimplification. There are substantial along-strike variations in the DSZ, including regions with only a single plane, merging planes, or cross-cutting features, as shown, for example, in Figure 7 of Münchmeyer et al. (2025). I’d assume you could find similar features in fine cross-sections of your catalog. Therefore, I’d suggest to use a more careful statement here.
Citation: https://doi.org/10.5194/essd-2025-254-RC2
Data sets
Automated seismicity catalogue and picks for the Chilean margin [Data set] M. Riedel-Hornig et al. https://doi.org/10.5281/zenodo.15284376
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,165 | 99 | 30 | 1,294 | 35 | 30 | 36 |
- HTML: 1,165
- PDF: 99
- XML: 30
- Total: 1,294
- Supplement: 35
- BibTeX: 30
- EndNote: 36
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This is an excellent manuscript and earthquake catalog for the years 2017 to 2021 spanning the Chilean subduction zone. This is a huge tectonic province with seismicity in the crust, on the plate boundary, and in the slab, and generation of a low completeness level, largely relocated in 3D structure, high quality catalog not plagued by network changes is a major contribution (of course, one hopes that the time span increases and there can be, say, five year updates going forward, as temporal patterns may be missed in the current catalog. The authors make many reasonable decisions of how to deal with the diverse sources of seismic signals and the vast along-strike length of the subduction zone, and these are clearly laid out. It seemed quite reasonable to include or exclude various data sources, and it is helpful to have it so clearly documented. The procedures applied to develop the catalog are widely utilized and appropriate for the catalog construction, with good description of the implementation of Earthquake Transformer, PyOcto, NonLinLoc and HypoDD, so the results are about the best that can be expected. The basic methods are not really unique or original, but the applications to the specific situation of Chile is. Some further lowering of completeness level might be possible for some subregions using template matching, but pushing the overall completeness down to magnitude 2.5 is already a major contribution and the catalog will be useful for many basic science applications. The catalog and phase picks are downloadable (already 106 downloads), and the map and cross-section figures show many intriguing features much more clearly than apparent in other catalogs.
The manuscript is very well written and the figures are excellent. I only caught one possible mistatement on line "135", where I believe "furthest" should be "closest", as duplicates are reduced by discarding the events in overlapping segments nearer to an edge. If that one statement is checked, I can recommend publication pretty much as is. This is a valuable contribution.