the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Oct 2024
Brief Communication: AI-driven rapid landslides mapping following the 2024 Hualien City Earthquake in Taiwan
Abstract. On April 2nd, 2024, a Mw 7.4 earthquake struck Taiwan’s eastern coast, triggering numerous landslides and severely impacting infrastructure. To create the preliminary inventory of earthquake-induced landslides in Eastern Taiwan (3,300 km2) we deployed automated landslide detection methods by combining Earth Observation (EO) data with Artificial Intelligence (AI) models. The models allowed us to identify 7,090 landslide events covering >75 km2, in about 3 hours after the acquisition of the EO imagery. This research underscores AI’s role in enhancing landslide detection for disaster response and situational awareness, and its implications for understanding earthquake-landslide interactions to improve seismic hazard mitigation.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Natural Hazards and Earth System Sciences.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.- Preprint
(3001 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on nhess-2024-146', Anonymous Referee #1, 31 Jan 2025
The authors presented a brief communication about the sinergic use of AI remote sensing tools, based on radar and optical imagery, and their implementation as a potential rapid mapping tool to be implemente for emergency response.
The paper is very interesting and promising in its goal, as well as well presented.
I have some minor comments which I hope will help its comprehension:
1) would you mind to improve the overall quality and readability of Figure 1? the study area boundaries and its topography are barely visible as it is.
2) Figure 2, it would be more effective to use the sum of the days after the triggering of the events in the timeline, instead of the dates.
3) line 25: what LMEs stand for?
4) you used the GDCLD database to train the model over Taiwan case of study. It would be more insightful to add some description of such database. Are these landslides comparable, in terms of geological geomorphological and land use setting, geometric and topographic features, to those triggered in Taiwan? Please, specify it.
5) You mentioned that 814 landslides were manually mapped. Are these landslides mapped in a specific area of Taiwan or allover the country? Did you focus on specific landslides or features?
6) I would like to see some lines about how to validate these landslides. Validation is completely missing here. Please, consider to add some remarks on how to validate landslides mapped by automatic tools.Citation: https://doi.org/10.5194/nhess-2024-146-RC1 -
RC2: 'Comment on nhess-2024-146', Anonymous Referee #2, 13 Feb 2025
The rapid extraction of landslides after the earthquake is an important task of post-earthquake relief and disaster assessment, and this manuscript carries out the post-earthquake landslide extraction of 2024 Hualien City Earthquake in Taiwan based on AI technology, which is an interesting work.
- The author said in the manuscript that no landslide inventory 15 for the 2024 Hualien City earthquake has been released, in fact, there are a number of articles on the landslide after the 2024 Hualien City Earthquake in Taiwan has been published.
2、page2: what DL and LMEs stand for?
- How to verify the accuracy of landslide extraction after the 2024 Hualien City Earthquake in Taiwan based on AI technology?
Citation: https://doi.org/10.5194/nhess-2024-146-RC2
Data sets
AI-driven rapid landslides mapping following the 2024 Hualien City Earthquake in Taiwan Lorenzo Nava, Alessandro Novellino, Chengyong Fang, Kushanav Bhuyan, Kathryn Leeming, Itahisa Gonzalez Alvarez, Claire Dashwood, Sophie Doward, Rahul Chahel, Emma McAllister, Sansar Raj Meena, Xuanmei Fan, Xiaochuan Tang, and Filippo Catani https://zenodo.org/records/11519683
Interactive computing environment
SAR-LRA Tool V1 for Google Colaboratory Lorenzo Nava, Alessandro Mondini, Kushanav Bhuyan, Oriol Monserrat, Alessandro Novellino, and Filippo Catani https://github.com/lorenzonava96/SAR-and-DL-for-Landslide-Rapid-Assessment/tree/main/SAR-LRA%20Tool%20V1%20for%20Google%20Colaboratory
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 447 | 119 | 135 | 701 | 16 | 18 |
- HTML: 447
- PDF: 119
- XML: 135
- Total: 701
- BibTeX: 16
- EndNote: 18
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1