A single framework to assess flash flood and landslide susceptibility: an application to the Mediterranean

Riveros, Alessia; Gunaratne, Chamidu; Martinelli, Mario; Sperna Weiland, Frederiek C.

doi:10.5194/egusphere-2025-5572

Preprints

https://doi.org/10.5194/egusphere-2025-5572

Preprints

04 Dec 2025

| 04 Dec 2025

A single framework to assess flash flood and landslide susceptibility: an application to the Mediterranean

Alessia Riveros, Chamidu Gunaratne, Mario Martinelli, and Frederiek C. Sperna Weiland

Abstract. Flash floods and landslides have caused severe economic damages and loss of life, especially in mountainous regions. To support effective risk management there is a growing interest in muti-hazard assessment. In this study a globally applicable Machine Learning (ML) Framework for landslide susceptibility mapping was extended to allow for the assessment of both landslide and flash flood susceptibility. The Framework was applied and evaluated in the Italian region Liguria that is frequently and severely impacted by both hazards. A relatively dense inventory of past events was constructed to facilitate the training of the ML Framework. The analysis revealed substantial similarities in the causative factors for the two hazards. There is a considerable area of Liguria susceptible to both hazards, although flash floods most often occur in river valleys whereas landslide susceptibility is also high in the upper courses of river catchments. We found a very high susceptibility along the coastline where many villages and cities are located. The unified framework allows for the integration of different hazard types under a consistent modelling structure. This enhances the comparability of results and supports the development of integrated mitigation strategies for any region of interest.

Received: 14 Nov 2025 – Discussion started: 04 Dec 2025

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 paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 1933 KB)

Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
Preprint (1933 KB)

Download & links

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Journal article(s) based on this preprint

01 Jun 2026

A single framework for assessing flash flood and landslide susceptibility: an application to the Mediterranean Liguria region, Italy

Alessia Riveros, Chamidu Gunaratne, Mario Martinelli, and Frederiek Christianne Sperna Weiland

Nat. Hazards Earth Syst. Sci., 26, 2437–2459, https://doi.org/10.5194/nhess-26-2437-2026,https://doi.org/10.5194/nhess-26-2437-2026, 2026

Short summary

Alessia Riveros, Chamidu Gunaratne, Mario Martinelli, and Frederiek C. Sperna Weiland

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2025-5572', Anonymous Referee #1, 23 Dec 2025

The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-5572/egusphere-2025-5572-RC1-supplement.pdf

Citation: https://doi.org/10.5194/egusphere-2025-5572-RC1
- AC1: 'Reply on RC1', Alessia Riveros, 27 Feb 2026
  
  Dear referee, thank you for your comments. Please find attached our responses
  
  Citation: https://doi.org/10.5194/egusphere-2025-5572-AC1
RC2:
'Comment on egusphere-2025-5572', Anonymous Referee #2, 07 Jan 2026
This manuscript investigates flash flood and landslide susceptibility using a unified machine-learning framework applied to the Liguria region in Italy. The study addresses an important topic in the context of multi-hazard risk assessment and benefits from the use of multiple event inventories, a consistent modelling framework across hazard types, and a careful discussion of several limitations. Overall, the paper is technically sound and clearly written, and it has the potential to make a valuable contribution to regional-scale multi-hazard susceptibility analysis. The comments below are intended to help strengthen the methodological clarity, interpretability, and positioning of the contribution.
The flash flood inventory spans a very long time period (approximately 1950 to present), whereas the susceptibility modelling relies on present-day static conditioning factors (e.g. land cover, river network, road networks, …). Over this period, substantial changes have occurred in Liguria, particularly related to urbanisation, river modification, and land-use change, which may violate the assumption of stationarity between past events and current susceptibility conditions. While the long inventory improves statistical robustness, it may also introduce temporal inconsistencies that affect the interpretation of the resulting susceptibility maps, especially in urbanised valleys where many early events predate current conditions. This issue is particularly relevant for flash floods, which are highly sensitive to anthropogenic hydrological modifications. The authors are encouraged to (i) acknowledge this limitation more explicitly, and/or (ii) demonstrate that it does not substantially affect the results. Possible approaches could include comparing susceptibility maps derived from a more recent subset of events (e.g. post-1990) with the full inventory-based map, or applying temporal weighting to events based on their distance from present-day conditions.

The Introduction states that this study transfers the framework of Tehrani et al. (2021) to flash flood susceptibility and applies it within a unified multi-hazard context. However, the precise methodological and conceptual advances beyond the original framework are not sufficiently articulated. The authors are encouraged to clarify more explicitly what aspects of the framework are novel (e.g. unified hazard treatment, inventory construction, comparative analysis), and how this work differs from or extends Tehrani et al. (2021) beyond a change in hazard type and study area. Strengthening this framing would help readers better appreciate the innovative potential of this study.

The title emphasises a Mediterranean-scale application, while the analysis is conducted at a local/regional scale focused on Liguria. While Liguria is representative of Mediterranean hydro-geomorphic settings, the current title may overstate the spatial scope of the application. The authors may wish to consider revising the title to better reflect the local-scale case study, while still acknowledging Mediterranean relevance.

In the Methodology section, the manuscript states that hyperparameters were calibrated using grid search and cross-validation, but the specific parameter ranges and final selections for each algorithm (LR, RF, SVM) are not described in detail. For transparency and reproducibility, the authors are encouraged to provide: the tested hyperparameter ranges, and the final selected values for each model, possibly in an appendix or supplementary material. This would strengthen the methodological clarity of the paper.

Class imbalance: the flash flood and landslide inventories differ substantially in size, and the binary classification problem is inherently unbalanced (event vs non-event pixels). While model performance metrics (AUC, accuracy, confusion matrices) are reported, the role of class imbalance in influencing false positives and false negatives is not explicitly discussed. The authors are encouraged to either: discuss how class imbalance may affect the reported performance metrics, or justify the decision not to apply balancing techniques (e.g. class weighting, SMOTE). Explicitly addressing this point would improve confidence in the interpretation of model performance.

To account for location uncertainty in point-based inventories, a 3x3 pixel window is used around each event. While this is a reasonable approach, it may introduce spatial smoothing that blurs the influence of certain conditioning factors, particularly for landslide susceptibility in steep terrain. As already partially acknowledged by the authors, this may contribute to unexpected results (e.g. landslides associated with gentler slopes). The authors may consider discussing alternative or complementary strategies, such as: exclusion or buffer zones around event pixels to reduce spatial autocorrelation, or sensitivity tests on window size. If considered too technical for implementation, a clearer discussion of the implications of this choice would still be valuable.

The framework is described as globally applicable; however, its performance depends strongly on the availability and quality of local hazard inventories. The authors are encouraged to expand the discussion on spatial and temporal transferability, particularly regarding:

data-rich versus data-scarce regions;

which conditioning variables may be problematic when transferring the model (e.g. the authors mention aspect -southward facing slopes- dependence, how would that change in regions with different dominant exposure?);

the potential role of high-resolution dynamic inputs (e.g. convection-permitting reanalysis and future projections) in improving applicability across regions and time periods.

This would help better position the framework for broader use beyond the Liguria case study.
Citation: https://doi.org/10.5194/egusphere-2025-5572-RC2
- AC2: 'Reply on RC2', Alessia Riveros, 27 Feb 2026
  
  Dear referee, thank you for your comments. Please find attached our responses.
  
  Citation: https://doi.org/10.5194/egusphere-2025-5572-AC2

Peer review completion

AR – Author's response | RR – Referee report | ED – Editor decision | EF – Editorial file upload

ED: Reconsider after major revisions (further review by editor and referees) (02 Mar 2026) by Timothy Tiggeloven

AR by Alessia Riveros on behalf of the Authors (10 Mar 2026) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (18 Mar 2026) by Timothy Tiggeloven

RR by Anonymous Referee #1 (18 Mar 2026)

RR by Davide Mauro Ferrario (14 Apr 2026)

Suggestions for revision or reasons for rejection

Comment 1: Frequency Ratio Analysis
The role of the Frequency Ratio (FR) analysis in the methodological framework is not sufficiently clear. Specifically, it remains ambiguous whether FR is employed solely as a descriptive tool to characterise the relationship between conditioning factors and the target phenomenon. The authors are encouraged to explicitly clarify the purpose of the FR analysis within the workflow.

Comment 2: Use of Accuracy as an evaluation metric
The use of Accuracy as a performance metric is problematic given the class imbalance present in the dataset. In skewed distributions, a classifier that systematically predicts the majority class can still achieve high accuracy while failing entirely on the minority class: for instance, in a dataset with 98% absence and 2% presence records, a trivial classifier predicting only absences would yield 98% accuracy. The authors are therefore encouraged to complement the AUC and confusion matrix with metrics better suited for imbalanced datasets, such as the F1-score, Revised Comment 2: Use of Accuracy as an Evaluation Metric
The use of Accuracy as a standalone performance metric is problematic in the context of imbalanced datasets, which are inherently common in landslide and natural hazard susceptibility studies, where landslide occurrences represent a small minority of the total observations. In such conditions, a model systematically predicting the majority class (absence) can still yield misleadingly high accuracy values while failing entirely on the minority class. The authors are encouraged to complement AUC with metrics better suited for imbalanced classification, such as the F1-score, Weighted Accuracy (i.e., class-weighted accuracy that accounts for class imbalance). See examples of use of F1 score for susceptibility mapping in Rusk et al., 2022 https://doi.org/10.1016/j.scitotenv.2021.150039; Tiggeloven et al., 2025: https://doi.org/10.1175/AIES-D-25-0039.1.

Comment 3: At line 407, an unresolved cross-reference appears ("Error: reference not found").

Hide

ED: Publish subject to minor revisions (review by editor) (15 Apr 2026) by Timothy Tiggeloven

AR by Alessia Riveros on behalf of the Authors (24 Apr 2026) Author's response Author's tracked changes Manuscript

ED: Publish as is (28 Apr 2026) by Timothy Tiggeloven

AR by Alessia Riveros on behalf of the Authors (12 May 2026) Manuscript

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2025-5572', Anonymous Referee #1, 23 Dec 2025

The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-5572/egusphere-2025-5572-RC1-supplement.pdf

Citation: https://doi.org/10.5194/egusphere-2025-5572-RC1
- AC1: 'Reply on RC1', Alessia Riveros, 27 Feb 2026
  
  Dear referee, thank you for your comments. Please find attached our responses
  
  Citation: https://doi.org/10.5194/egusphere-2025-5572-AC1
RC2:
'Comment on egusphere-2025-5572', Anonymous Referee #2, 07 Jan 2026
This manuscript investigates flash flood and landslide susceptibility using a unified machine-learning framework applied to the Liguria region in Italy. The study addresses an important topic in the context of multi-hazard risk assessment and benefits from the use of multiple event inventories, a consistent modelling framework across hazard types, and a careful discussion of several limitations. Overall, the paper is technically sound and clearly written, and it has the potential to make a valuable contribution to regional-scale multi-hazard susceptibility analysis. The comments below are intended to help strengthen the methodological clarity, interpretability, and positioning of the contribution.
The flash flood inventory spans a very long time period (approximately 1950 to present), whereas the susceptibility modelling relies on present-day static conditioning factors (e.g. land cover, river network, road networks, …). Over this period, substantial changes have occurred in Liguria, particularly related to urbanisation, river modification, and land-use change, which may violate the assumption of stationarity between past events and current susceptibility conditions. While the long inventory improves statistical robustness, it may also introduce temporal inconsistencies that affect the interpretation of the resulting susceptibility maps, especially in urbanised valleys where many early events predate current conditions. This issue is particularly relevant for flash floods, which are highly sensitive to anthropogenic hydrological modifications. The authors are encouraged to (i) acknowledge this limitation more explicitly, and/or (ii) demonstrate that it does not substantially affect the results. Possible approaches could include comparing susceptibility maps derived from a more recent subset of events (e.g. post-1990) with the full inventory-based map, or applying temporal weighting to events based on their distance from present-day conditions.

The Introduction states that this study transfers the framework of Tehrani et al. (2021) to flash flood susceptibility and applies it within a unified multi-hazard context. However, the precise methodological and conceptual advances beyond the original framework are not sufficiently articulated. The authors are encouraged to clarify more explicitly what aspects of the framework are novel (e.g. unified hazard treatment, inventory construction, comparative analysis), and how this work differs from or extends Tehrani et al. (2021) beyond a change in hazard type and study area. Strengthening this framing would help readers better appreciate the innovative potential of this study.

The title emphasises a Mediterranean-scale application, while the analysis is conducted at a local/regional scale focused on Liguria. While Liguria is representative of Mediterranean hydro-geomorphic settings, the current title may overstate the spatial scope of the application. The authors may wish to consider revising the title to better reflect the local-scale case study, while still acknowledging Mediterranean relevance.

In the Methodology section, the manuscript states that hyperparameters were calibrated using grid search and cross-validation, but the specific parameter ranges and final selections for each algorithm (LR, RF, SVM) are not described in detail. For transparency and reproducibility, the authors are encouraged to provide: the tested hyperparameter ranges, and the final selected values for each model, possibly in an appendix or supplementary material. This would strengthen the methodological clarity of the paper.

Class imbalance: the flash flood and landslide inventories differ substantially in size, and the binary classification problem is inherently unbalanced (event vs non-event pixels). While model performance metrics (AUC, accuracy, confusion matrices) are reported, the role of class imbalance in influencing false positives and false negatives is not explicitly discussed. The authors are encouraged to either: discuss how class imbalance may affect the reported performance metrics, or justify the decision not to apply balancing techniques (e.g. class weighting, SMOTE). Explicitly addressing this point would improve confidence in the interpretation of model performance.

To account for location uncertainty in point-based inventories, a 3x3 pixel window is used around each event. While this is a reasonable approach, it may introduce spatial smoothing that blurs the influence of certain conditioning factors, particularly for landslide susceptibility in steep terrain. As already partially acknowledged by the authors, this may contribute to unexpected results (e.g. landslides associated with gentler slopes). The authors may consider discussing alternative or complementary strategies, such as: exclusion or buffer zones around event pixels to reduce spatial autocorrelation, or sensitivity tests on window size. If considered too technical for implementation, a clearer discussion of the implications of this choice would still be valuable.

The framework is described as globally applicable; however, its performance depends strongly on the availability and quality of local hazard inventories. The authors are encouraged to expand the discussion on spatial and temporal transferability, particularly regarding:

data-rich versus data-scarce regions;

which conditioning variables may be problematic when transferring the model (e.g. the authors mention aspect -southward facing slopes- dependence, how would that change in regions with different dominant exposure?);

the potential role of high-resolution dynamic inputs (e.g. convection-permitting reanalysis and future projections) in improving applicability across regions and time periods.

This would help better position the framework for broader use beyond the Liguria case study.
Citation: https://doi.org/10.5194/egusphere-2025-5572-RC2
- AC2: 'Reply on RC2', Alessia Riveros, 27 Feb 2026
  
  Dear referee, thank you for your comments. Please find attached our responses.
  
  Citation: https://doi.org/10.5194/egusphere-2025-5572-AC2

Peer review completion

AR – Author's response | RR – Referee report | ED – Editor decision | EF – Editorial file upload

ED: Reconsider after major revisions (further review by editor and referees) (02 Mar 2026) by Timothy Tiggeloven

AR by Alessia Riveros on behalf of the Authors (10 Mar 2026) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (18 Mar 2026) by Timothy Tiggeloven

RR by Anonymous Referee #1 (18 Mar 2026)

RR by Davide Mauro Ferrario (14 Apr 2026)

Suggestions for revision or reasons for rejection

Hide

ED: Publish subject to minor revisions (review by editor) (15 Apr 2026) by Timothy Tiggeloven

AR by Alessia Riveros on behalf of the Authors (24 Apr 2026) Author's response Author's tracked changes Manuscript

ED: Publish as is (28 Apr 2026) by Timothy Tiggeloven

AR by Alessia Riveros on behalf of the Authors (12 May 2026) Manuscript

Journal article(s) based on this preprint

01 Jun 2026

A single framework for assessing flash flood and landslide susceptibility: an application to the Mediterranean Liguria region, Italy

Alessia Riveros, Chamidu Gunaratne, Mario Martinelli, and Frederiek Christianne Sperna Weiland

Nat. Hazards Earth Syst. Sci., 26, 2437–2459, https://doi.org/10.5194/nhess-26-2437-2026,https://doi.org/10.5194/nhess-26-2437-2026, 2026

Short summary

Alessia Riveros, Chamidu Gunaratne, Mario Martinelli, and Frederiek C. Sperna Weiland

Model code and software

LHAT Robyn Gwee et al. https://doi.org/10.5281/zenodo.17579993

Alessia Riveros, Chamidu Gunaratne, Mario Martinelli, and Frederiek C. Sperna Weiland

Viewed

Total article views: 2,959 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
1,753	1,029	177	2,959	146	148

HTML: 1,753
PDF: 1,029
XML: 177
Total: 2,959
BibTeX: 146
EndNote: 148

Views and downloads (calculated since 04 Dec 2025)

Month	HTML	PDF	XML	Total
Dec 2025	720	364	85	1,169
Jan 2026	465	180	40	685
Feb 2026	241	171	23	435
Mar 2026	206	154	15	375
Apr 2026	53	55	5	113
May 2026	18	60	1	79
Jun 2026	8	3	0	11
Jul 2026	4	15	2	21
Aug 2026	6	7	2	15
Sep 2026	27	15	2	44
Oct 2026	5	5	2	12

Cumulative views and downloads (calculated since 04 Dec 2025)

Month	HTML	PDF	XML	Total
Dec 2025	720	364	85	1,169
Jan 2026	465	180	40	685
Feb 2026	241	171	23	435
Mar 2026	206	154	15	375
Apr 2026	53	55	5	113
May 2026	18	60	1	79
Jun 2026	8	3	0	11
Jul 2026	4	15	2	21
Aug 2026	6	7	2	15
Sep 2026	27	15	2	44
Oct 2026	5	5	2	12

Viewed (geographical distribution)

Total article views: 2,927 (including HTML, PDF, and XML) Thereof 2,927 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 06 Oct 2026

Download

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Preprint (1933 KB)
Metadata XML

Short summary

Flash floods and landslides have caused severe economic damages and loss of life, especially in mountainous regions e.g. Liguria. We created susceptibility maps to both hazards based on past recorded events and open-data such as slopes and altitude. We found a similar high predisposition to both hazards along the coast. Outside the coastal area, river valleys and urban areas (or upper river courses) exhibited high susceptibility only to flash floods (or landslides).


Total:	0
HTML:	0
PDF:	0
XML:	0