the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Comparison of conditioning factors classification criteria in large scale statistically based landslide susceptibility models
Abstract. The large scale landslide susceptibility assessment (LSA) is an important tool for reducing landslide risk through the application of resulting maps in spatial and urban planning. The existing literature more often deals with LSA modelling techniques and the scientific research very rarely focuses on acquiring relevant thematic and landslide data, necessary to achieve reliable results. Therefore, the paper focuses on the crucial step of classifying continuous landslide conditioning factors for susceptibility modelling by presenting an innovative comprehensive analysis that resulted in 54 landslide susceptibility models to test 11 classification criteria (scenarios which vary from stretched values, partially stretched classes, heuristic approach, classification based on studentized contrast and landslide presence, and commonly used classification criteria, such as Natural Neighbor, Quantiles and Geometrical intervals) in combination with five statistical methods. The large scale landslide susceptibility models were derived for small and shallow landslides in the pilot area (21 km2) located in the City of Zagreb (Croatia), which occur mainly in soils and soft rocks. Some of the novelties in LSA are the following: scenarios using stretched landslide conditioning factor values or classification with more than 10 classes prove more reliable; certain statistical methods are more sensitive to the landslide conditioning factor classification criteria than others; all the tested machine learning methods give the best landslide susceptibility model performance using continuously stretched landslide conditioning factors derived from high-resolution input data. The research highlights the importance of qualitative assessments, alongside commonly used quantitative metrics, to verify spatial accuracy and to test the applicability of derived landslide susceptibility maps for spatial planning purposes.
- Preprint
(2932 KB) - Metadata XML
-
Supplement
(24108 KB) - BibTeX
- EndNote
Status: open (until 22 May 2024)
-
RC1: 'Comment on nhess-2024-29', Anonymous Referee #1, 27 Mar 2024
reply
Dear Editor,
Please find below my review of the paper nhess-2024-29:
Comparison of conditioning factor classification criteria in large scale statistically based landslide susceptibility models
by
Marko Sinčić, Sanja Bernat Gazibara, Mauro Rossi, Snježana Mihalić Arbanas
This paper is related to the comparison of landslide susceptibility models. It uses landslide conditioning factors using different classification scenarios using 5 models. Results are discussed.
General Comments
The paper is well written, and references are appropriated, it is a valuable contribution to susceptibility model assessment, but a few problems remain. The paper is an attempt to objectivise comparison of methods and the impact of classification of conditioning factors. If from page 14 things are clear, it is more difficult to follow the pages 5 to 13. A clear description of the scenario’s meaning is necessary as well. The first clear sentence about the relationship between several important elements appears in lines 271-273: “Furthermore, 54 LSMs were derived using the prepared landslide dataset and 11 LCF sets in the five selected methods, i.e., Information Value (IV), Logistic Regression (LR), Neural Network (NN), Random Forests (RF) and Support Vector Machine (SVM).”, which must appear really earlier, but even though I count only 7 LCF in table 3 (this discrepancy must at least be introduced in the table legend).
I am sure that all the information, but for the reader (at least for me) it is difficult to go in the paper. I would recommend introducing a figure of a graphical flow chart to explain the relationships between LCF, LSM and section 3.2, figure 2 being rather difficult to follow (maybe adding image may help). This will greatly improve the understanding of the paper.
I also recommend carefully rereading sections 4.2 to 4.6 regarding discussions, there are some redundancies that can be avoided by having some arguments in the discussion, since the paper is already rather long.
Few specific comments
- Don does not forget parentheses for dates of publications.
- Corominas et al. is 2014 not 2013 (online version)
- Figures 8 and 9, if I understand well the first is linked area and the second is related on pixels.
Citation: https://doi.org/10.5194/nhess-2024-29-RC1
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
298 | 35 | 16 | 349 | 19 | 12 | 13 |
- HTML: 298
- PDF: 35
- XML: 16
- Total: 349
- Supplement: 19
- BibTeX: 12
- EndNote: 13
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1