Next Article in Journal
Simulation of a Multi-Stage Stress Field and Regional Prediction of Structural Fractures in the Tucheng Syncline, Western Guizhou, China
Previous Article in Journal
Landslide at the River’s Edge: Alum Bluff, Apalachicola River, Florida
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Geosciences
Volume 15
Issue 4
10.3390/geosciences15040131
geosciences-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Landslides in the Himalayas: The Role of Conditioning Factors and Their Resolution in Susceptibility Mapping

by
Lalit Pathak
1,
Badri Baral
2,3,
Kamana Joshi
4,
Dipak Raj Basnet
2,3 and
Danilo Godone
5,*
1
Centre of Research for Environment, Energy and Water (CREEW), Kathmandu 44616, Nepal
2
Nature Conservation Initiative Nepal (NCI-Nepal), Gokarneshwor-05, Kathmandu 44602, Nepal
3
Nepal Environmental Research Institute (NERI), Tarakeshwor-09, Kathmandu 44610, Nepal
4
Central Department of Environmental Science (CDES), Tribhuvan University, Kathmandu 44613, Nepal
5
Research Institute for Geo-Hydrological Protection (CNR-IRPI) Turin, National Research Council of Italy, 10135 Torino, Italy
*
Author to whom correspondence should be addressed.
Geosciences 2025, 15(4), 131; https://doi.org/10.3390/geosciences15040131
Submission received: 14 February 2025 / Revised: 20 March 2025 / Accepted: 27 March 2025 / Published: 2 April 2025
(This article belongs to the Section Natural Hazards)
Browse Figures
Versions Notes

Abstract

:
Landslides present remarkable hazards in the Himalayan region, particularly in areas with young and fragile topography. Mitigating vulnerability requires assessing susceptibility, which relies heavily on the accuracy of susceptibility maps generated through various approaches that consider different conditioning factors at various resolutions. This study, conducted in Jajarkot District within the Karnali Province of Nepal and covering 2230 km2, aims to identify suitable conditioning factors at appropriate resolutions. Sixteen factors, encompassing topography, hydrology, geology, and anthropogenic activities, were analyzed alongside a landslide inventory of 159 occurrences compiled from satellite imagery, the literature, and field surveys. A genetic algorithm (GA) was employed to determine the optimal set of conditioning factors, while Maximum Entropy (Maxent) modeling produced landslide susceptibility maps (LSM) at spatial resolutions ranging between 12.5 and 200 m. Resolution selection was guided by Receiver Operating Characteristic (ROC) curve and Area Under the Curve (AUC) analyses. Multicollinearity testing identified 15 influential factors, with land use ranking highest at 22.7%, followed by stream power index (SPI), drainage density, and aspect. The GA consistently highlighted land use and slope as effective factors across subset sizes. The results indicated resolutions finer than one hundred meters enhanced discrimination between landslide and non-landslide areas, emphasizing the need to balance resolution with computational resources and data availability. This study emphasizes the intricate interplay of conditioning factors, the GA’s efficacy in subset selection, and the crucial role of resolution in the improvement of susceptibility models. The findings provide practical insights for policymakers and disaster management authorities, aiding evidence-based decision making in the mitigation of landslide risk in Jajarkot and similar regions.

1. Introduction

Landslides are noteworthy calamitous events, leading to property damage, loss of life, and adverse impacts on biodiversity in the Himalayas [1,2,3]. The combination of dynamic tectonic activity, rugged terrain, steep slopes, ongoing geological processes, and anthropic activities (e.g., deforestation and haphazard infrastructure development), renders this region highly susceptible to landslides [4,5]. Consequently, there is an escalation in interest in the development of accurate and reliable landslide susceptibility maps (LSMs) for the area to facilitate targeted mitigation strategies. Allied with this, there is considerable literature on landslide susceptibility modeling, an approach that infers the likelihood of a landslide in any region [6,7,8,9,10,11,12,13]. Yet, the complex nature of landslides and their complicated interactions with conditioning factors make it difficult to identify suitable conditioning factors and landslide-susceptible zones [9,10,14,15].
Preparation of the conditioning factors necessary for susceptibility analysis is inherently difficult [16], with no universal rule regarding the optimal quantity of factors for analysis. It is unclear which particular landslide susceptibility problem is best solved by which particular combination of parameters because of the range of landslide-related characteristics [17,18,19]. Additionally, it is quite likely that correlated and redundant data of all available parameters are likely to lower the accuracy of the final map [20,21]. Refs. [22,23,24,25,26] examined a limited number of factors, whereas Ref. [27] explored multiple combinations of factors and chose the most effective ones. Similarly, some studies also included landslide-triggering factors such as rainfall [1,9] and earthquakes [25,28]. In the Himalayan region, the most frequently used conditioning factors included are aspect, elevation, profile curvature, plane curvature, slope, topographic positioning index (TPI), proximity to rivers, drainage density, stream power index (SPI), topographic wetness index (TWI), proximity to roads, land use, proximity to thrust, dominant soils, and lithology [3,10,29,30].
At the same time, the complexity of incorporating numerous landforms and spatial information poses challenges in achieving data and methodological consistency [31]. Furthermore, a standardized framework for selecting these factors is lacking, and often relies on users’ experience and experts’ opinions for guidance [15,32]. In this paucity of specific approaches to the selection of conditioning factors for landslides, the genetic algorithm (GA) can be implemented to identify the optimal combination of landslide conditioning factors [33,34,35,36]. The GA has been used extensively in factor optimization and to identify the most suitable subset of conditioning factors [27]. Along with the selection of conditioning factors, the resolution of data used also plays a crucial role in the accuracy and reliability of landslide susceptibility mapping [35,37,38].
High-resolution data on topography, geology, and land cover are crucial in the accurate identification of areas prone to landslides; however, high-resolution data may not always provide a detailed representation of landslide-prone zones [35,39]. Maxent is a versatile machine learning model designed for presence-only data analysis [15,40]. Initially employed in species habitat suitability assessment, scientists have started utilizing this model to evaluate landslide susceptibility [41]. The results of these evaluations have proven effective and superior to models like support vector machines and artificial neural networks [21,24,42]. Additionally, the fundamental theory of landslide assessment suggests that future landslides are likely to occur in areas with a history of landslides [43]. This model utilizes Bayesian inference to quantitatively assess geographical probability by considering geological and environmental variables associated with presence data [44].
In this study, we investigated the role of conditioning factors and their resolution in landslide susceptibility mapping to provide a thematic map that portrays where landslides are likely to occur and the probability of their occurrence in Jajarkot within the Himalayan region. The aim was to contribute to a better understanding of factors influencing landslide susceptibility and provide insight into methods that improve landslide hazard assessment and management in the region.

2. Materials and Methods

2.1. Study Area

Jajarkot located in Karnali Province, with mountain specificities in the Western Himalayan region of Nepal spans an area of about 2230 km2 and is geographically positioned between 81°49.19′ and 82°34.80′ E longitude and 28°36.67′ and 29°7.64′ N latitude (Figure 1). The temperature varies between 8 °C and 34 °C, with an average annual rainfall of 1868.5 mm. The study area extends from 672 masl. to 5384 masl. and is situated between the lesser Himalayan and higher Himalayan regions, with the Main Central Thrust (MCT) traversing through it [45]. The study area beholds sub-tropical, temperate, sub-alpine, and nival types of vegetation communities [46]. The topography is characterized by a rugged and extensively dissected landscape. Given its topography and climatic conditions, Jajarkot is highly susceptible to landslides and poses a high risk to the people living in the area.

2.2. Methods

2.2.1. Landslide Inventory Mapping

A comprehensive landslide inventory was developed using satellite imagery and field surveys following Ref. [3]. Utilizing various satellite sources (Google Earth Pro time series images, Landsat, and Sentinel imagery) and past studies Refs. [47,48], 195 landslides were identified and digitized in QGIS and Google Earth Pro. Field validation confirmed 50 percent of the landslides to ensure accuracy. A photograph depicting a landslide observed during field verification is presented in Figure 2. Centroid coordinates were extracted for Maxent modeling, with 70 percent of landslides used for training and 30 percent for validation, aligning with the approach by Ref. [49].

2.2.2. Conditioning Factors (CFs)

Based on regional environmental characteristics of the study area and the scientific literature, 16 landslide conditioning factors were selected, as depicted in Figure 3. The data were acquired from various sources to derive the physical, hydrological, geological, and anthropic conditioning factor layers. Physical conditioning factors were derived using a Digital Elevation Model (DEM) acquired from Alaska Satellite facility (http://asf.alaska.edu#homepagehttp://asf.alaska.edu#homepage, accessed on 5 January 2025), following the methodologies suggested in Refs. [50,51,52]. Geological factors and lithological maps were acquired from ICIMOD and soil data from the SOTER database were used. Similarly, data on land use, which plays a key role in the operation of hydrological and geomorphological processes either directly or indirectly, were acquired from ESRI Global land cover, 2020 [53]. The conditioning factors were then resampled to a set of (12.5 m, 25 m, 50 m, 100 m and 200 m) pixels for further analysis. All geospatial analyses were performed using QGIS Desktop (version 3.28.9).

2.2.3. Multicollinearity Test

Multicollinearity among conditioning factors was estimated by means of a Variation Inflation Factor (VIF) to reduce errors in the analyses due to high correlation between conditioning factors [54,55]. Among the 15 conditioning factors (Figure 4), the multicollinearity test, as suggested by Refs. [56,57] was implemented using the JupyterLab (version 4.0.0) interface.

2.2.4. Optimal Subsets of Conditioning Factors

The GA was implemented to identify the optimal combination of landslide conditioning factors, as the selection of optimal subsets is crucial in landslide susceptibility mapping [37,38]. In the current study, 15 available conditioning factors were considered to investigate the most effective subset of factors following Ref. [27]. Mahalanobis distance was used as a fitness function to evaluate the quality of each combination (represented by chromosomes in the GA) and a fitness score was calculated for 12 different subset sizes (ranging between 4 and 15) that represents the statistical measure of distinction between landslide and non-landslide samples. Each pixel in the test dataset was then assessed using the Mahalanobis distance and assigned a label according to the closest centroid [27].

2.2.5. Maximum Entropy Model for Landslide Susceptibility

The Maxent model was used in the landslide susceptibility assessment using a point-based landslide inventory and conditioning factors at different resolutions. The fundamental evaluation principle involves dividing the study area into a set of pixels represented by the variable X, where each computing unit x ∈ X denotes a specific location. The probability distribution value of each occurrence point within each computing unit, denoted by π(x), is assigned a value between 0 and 1. The sum of the probability values across all computing units is equal to 1. The response variable y is represented by the presence value in each calculation grid, where y = 1 indicates occurrence of a landslide and 0 represents its absence. Consequently, the probability of occurrence, based on the conditional distribution P(y|x), is as follows:
P (y = 1|x) = [P(x(x|y = 1) P(y = 1)]/P(x),
where P (y = 1|x) is the probability of occurrence at a specific point x; P(y|x = 1) is the probability of a landslide at a specific point x under presence points-based distribution conditions, which is also π(x); P(y = 1) is the overall incidence of landslides; and P(x) is the probability of landslide occurrence at any point x. During the model’s calibration phase, a random selection algorithm was employed to train the model, with 70% of the datasets randomly chosen, output was gained and set as ASCII, and Jackknife analysis was performed.

2.2.6. Assessment of the Optimal Resolution

The optimal resolution for the landslide susceptibility model was investigated by evaluating it at a 12.5 m, 25 m, 50 m, 100 m, and 200 m resolution by resampling conditioning factors; the same test and training data were used for consistency. The model outputs were then validated at each resolution to assess their performance. Evaluation of the model’s output was performed using a ROC curve, a systematic method used to assess the accuracy of deterministic and probabilistic prediction systems. In this approach, the area under the curve (AUC) was used as a qualitative measure of prediction accuracy, which has been categorized into three classes. An AUC value between 0.7 and 0.8 indicates acceptable performance; between 0.8 and 0.9 indicates excellent performance; and a value above 0.9 indicates outstanding performance [58].

3. Results

3.1. Factor Assignment and Importance

The result of factor importance, in terms of percent contribution and permutation importance (Figure 5), greatly impacts the preparation of a susceptibility map [41,59]. Among the CFs, land use demonstrated the highest percentage contribution, with a value of 22.7%, indicating its strong influence on landslide susceptibility. Other selected CFs, such as slope, TPI, dominant soil, proximity to thrust, and elevation, exhibited lower percent contributions, ranging between 0.2% and 3.5% (Figure 5).

3.2. CF Selection

One of the crucial processes in susceptibility assessment is the multicollinearity test, which helps to determine whether predisposing factors are redundant in relation to other factors and remove them [60]. The results of the multicollinearity test with VIF values are illustrated in Table 1. Many landslide conditioning factors have an influence on landslide susceptibility mapping, but it is not certain which factors produce the best results for a given area under analysis. With the availability of an increasing number of landslide conditioning factors, finding the best combination of factors has become an important research issue. When the tolerance value is <0.1 or the VIF is >10, it indicates the existence of multicollinearity among the conditioning factors [57,61]. In our analysis, the VIF value of the TRI is 15.09, which is >10, indicating its removal; the results after removal of the TRI show no multicollinearity among the conditioning factors, therefore, the 15 factors tested were used in the susceptibility analysis.
Fitness scores, calculated by the Mahalanobis distance for the different subset sizes, indicate the statistical measure of separability between landslide and non-landslide samples. Using a validation dataset, the highest score achieved was for a 12-subset solution and the lowest score was for a 4-subset solution (Table 2).

3.3. Best Performing Resolution

Validation of each susceptibility map (Figure 6) at different resolutions was assessed to identify the best performing resolution of the datasets using ROC curve analysis [10,39]. At a resolution of 12.5 m, the model reached a training and test AUC of 0.867 and 0.828, respectively. Increasing the resolution to 25 m resulted in improved performance, with a training and test AUC of 0.889 and 0.862, respectively (Figure 7). This enhancement trend continued as the resolution increased to 50 m, yielding a training and test AUC of 0.908 and 0.864, respectively. At a resolution of 100 m, the model further improved, reaching a training and test AUC of 0.920 and 0.878, respectively.

4. Discussion

Land use plays a significant role in the stability of the slope and flow of water, as well as infiltration, and is the most widely used landslide condition factor [62]. Our results suggest that land use has a relatively influential impact on the improvement of the accuracy of LSMs, also demonstrated by Ref. [62]. SPI, drainage density, and aspect also make substantial contributions, with percentage contributions of 17.3%, 15.8%, and 10.8%, respectively. Drainage density is an important factor in many hazards, including landslides. Landslides are caused by slope failure and a high drainage density can increase the risk of a landslide by increasing the amount of surface runoff. Surface runoff is the water that flows over the surface of the ground instead of infiltrating into the soil. When there is a lot of surface runoff, it can increase the amount of water available to erode the soil [63]. This can make the soil more likely to fail and cause a landslide [39]. Similarly, when the SPI is high, there is more potential energy available to increase the amount of sediment, which means that there is greater potential for erosion and increases the probability of a landslide [64]. Although factors such as slope, TPI, dominant soil, proximity to thrust, and elevation have been widely considered due to their potential to influence landslide susceptibility [41], slope angle is also considered a direct predictor of landslide occurrence [22]. It reflects the steepness of each grid surface, which contributes much to slope stability. Generally, steeper slopes are likely to experience more landslides. In the case of the TPI, it is not considered a perfect predictor of landslide susceptibility [65]. In terms of permutation importance, drainage density emerges as the most important factor, with a value of 25.4%, followed by dominant soil with a value of 14.3%, and lithology with a value of 10.2%. These findings provide valuable insight into the relative importance of different factors in susceptibility mapping, guide future landslide susceptibility assessment, and aid in the comprehension and management of landslide-prone areas [41,66,67].
From the GA, it is clear that land use and slope are the most effective factors; as they were selected in almost all subsets. In addition, lithology, proximity to roads, and elevation were frequently selected. Similar types of results are suggested by Ref. [10], where slope, slope aspect, geology, and road construction activity (an anthropogenic cause) are suggested as the most effective predisposing causative parameters of landslide activity in the Nepal Himalayan region.
Validation of each susceptibility map, as demonstrated by Ref. [35], indicates that at the highest resolution considered, 200*200, there is a slight decrease in performance, with respective training and test AUCs of 0.889 and 0.861. Therefore, the results suggest that it is important to balance resolution with computational resources and data availability [35,39]. Tian et al. [35] also conclude that finer resolution does not necessarily lead to higher accuracy in landslide susceptibility mappings; they found a “W” shape accuracy curve in their sensitivity analysis, indicating a 90-m resolution to be the most accurate and a 150-m resolution to be least. Similarly, Meena et al. [39] suggest a 30-m resolution and Chen et al. [68] suggest 70-m resolution maps as the best contributing factor in their analysis, suggesting that moderate resolutions may better capture the geomorphological features relevant to landslide occurrence. Similarly, Claessens et al. [69] observe that finer DEMs tend to overestimate stable areas due to an increase in cell count, leading to reduced landslide probabilities, while coarser resolutions may smooth out critical terrain variations. Catani et al. [38] also report that optimal susceptibility maps were obtained using resolutions between 50 m and 100 m, which is consistent with our observation that an intermediate resolution provides the best trade-off. Collectively, these studies emphasize that the selection of DEM resolution should be guided by the characteristic landslide scale, regional topographic complexity, and computational resources available. The impact of spatial resolution on the accuracy of landslide susceptibility mapping is evident; however, their relationship is not strictly linear. The tendency to adopt increasingly finer grid cells lacks clear justification. Instead, the optimal resolution should be determined by carefully considering mapping accuracy, data availability, and the representative scale of landslides.

5. Conclusions

This study contributes to the existing knowledge by employing a comprehensive approach that integrates detailed landslide inventory mapping, careful conditioning factor selection, and advanced modeling techniques which will remain crucial in dealing with the vulnerability and risk assessment aftermath of the 2023 earthquake in Jajarkot. This study provides a thematic map that portrays the likelihood of a landslide to occur there in order to mitigate risk and take preventative action during reconstruction.
Land use emerged as the most influential conditioning factor, underscoring the impact of anthropogenic activities on susceptibility patterns. The incorporation of these factors, guided by the GA, enhances the accuracy and reliability of the Maxent model, offering a nuanced understanding of the multifaceted nature of landslides. The GA consistently highlighted the importance of land use and slope in the various subsets. Lithology, proximity to roads, and elevation also featured prominently, emphasizing their relevance when modeling landslide susceptibility. This iterative selection process aids in streamlining future assessment and provides a foundation for targeted mitigation strategies.
This study contributes to ongoing discourse on the role of resolution in landslide hazard assessment, emphasizes the importance of balancing resolution considerations with computational resources and data availability, and offers practical guidance for future studies in similar contexts. The identification of influential conditioning factors, the utilization of advanced modeling techniques, and the exploration of optimal resolutions contribute valuable insights that can inform evidence-based decision making in landslide-prone areas via LSM application. These maps can serve as crucial tools for policymakers, land use planners, and disaster management authorities, aiding in the development of targeted strategies for risk reduction and resilience building. Future research should focus on integrating emerging technologies into landslide susceptibility models to enhance their accuracy and timeliness. For example, near-real-time remote sensing data, such as high-frequency SAR imagery from Sentinel-1 or high-resolution optical data from Sentinel-2, and commercial satellites can be used to monitor ground deformation and precipitation patterns that often precede landslide events. Additionally, incorporating climate change projections (e.g., using RCP scenarios) into model inputs can help assess how shifts in precipitation patterns, temperature, and extreme weather events might alter landslide hazards over time. Incorporating dynamic land use data from UAVs or platforms like Google Earth Engine can help capture changes in vegetation and urbanization. Combining these datasets with machine learning algorithms and artificial intelligence could lead to adaptive, time-variant models that not only predict static susceptibility but also forecast changes in landslide hazards under different future scenarios. Furthermore, validating these adaptive models across diverse Himalayan regions will further refine their accuracy and generalizability.

Author Contributions

Conceptualization, L.P., B.B., K.J. and D.R.B.; data curation, L.P., B.B., K.J., D.R.B. and D.G.; formal analysis, L.P., B.B., K.J., D.R.B. and D.G.; methodology, L.P., B.B. and D.R.B.; writing—original draft, L.P., B.B., K.J. and D.R.B.; writing—review and editing, L.P., B.B., D.R.B. and D.G. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

Dataset available on request from the authors.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Dikshit, A.; Sarkar, R.; Pradhan, B.; Segoni, S.; Alamri, A.M. Rainfall Induced Landslide Studies in Indian Himalayan Region: A Critical Review. Appl. Sci. 2020, 10, 2466. [Google Scholar] [CrossRef]
  2. Mirdda, H.A.; Bera, S.; Chatterjee, R. Vulnerability Assessment of Mountainous Households to Landslides: A Multidimensional Study in the Rural Himalayas. Int. J. Disaster Risk Reduct. 2022, 71, 102809. [Google Scholar] [CrossRef]
  3. Pathak, L.; Devkota, K.C. Distribution and Dynamic Behaviors of Landslide in Rangun Khola Watershed of the Western Nepal. J. Environ. Sci. 2022, VIII, 68–79. [Google Scholar] [CrossRef]
  4. Dhakal, S. Geological Divisions and Associated Hazards in Nepal. In Contemporary Environmental Issues and Methods in Nepal; Springer: Berlin/Heidelberg, Germany, 2014; pp. 100–109. [Google Scholar]
  5. Dhital, M. Geology of the Nepal Himalaya: Regional Perspective of the Classic Collided Orogen; Springer: Berlin/Heidelberg, Germany, 2015; ISBN 978-3-319-02495-0. [Google Scholar]
  6. Brabb, E.E. Innovative Approaches to Landslide Hazard and Risk Mapping. In Proceedings of the International Landslide Symposium Proceedings, Toronto, ON, Canada, 23–31 August 1985; Volume 1, pp. 17–22. [Google Scholar]
  7. Chung, C.J.F.; Fabbri, A.G. Probabilistic Prediction Models for Landslide Hazard Mapping. Photogramm. Eng. Remote Sens. 1999, 65, 1389–1399. [Google Scholar]
  8. Cignetti, M.; Godone, D.; Bertolo, D.; Paganone, M.; Thuegaz, P.; Giordan, D. Rockfall Susceptibility along the Regional Road Network of Aosta Valley Region (Northwestern Italy). J. Maps 2021, 17, 54–64. [Google Scholar] [CrossRef]
  9. Dahal, R.K.; Hasegawa, S. Representative Rainfall Thresholds for Landslides in the Nepal Himalaya. Geomorphology 2008, 100, 429–443. [Google Scholar] [CrossRef]
  10. Devkota, K.C.; Regmi, A.D.; Pourghasemi, H.R.; Yoshida, K.; Pradhan, B.; Ryu, I.C.; Dhital, M.R.; Althuwaynee, O.F. Landslide Susceptibility Mapping Using Certainty Factor, Index of Entropy and Logistic Regression Models in GIS and Their Comparison at Mugling-Narayanghat Road Section in Nepal Himalaya. Nat. Hazards 2012, 65, 135–165. [Google Scholar] [CrossRef]
  11. Guzzetti, F.; Cardinali, M.; Reichenbach, P.; Carrara, A. Comparing Landslide Maps: A Case Study in the Upper Tiber River Basin, Central Italy. Environ. Manag. 2000, 25, 247–263. [Google Scholar] [CrossRef]
  12. Pathak, L.; Devkota, K.C. Landslide Susceptibility Assessment in the Rangun Khola Watershed of Far Western Nepal. J. Nepal. Geol. Soc. 2022, 63, 33–44. [Google Scholar] [CrossRef]
  13. Regmi, N.R.; Giardino, J.R.; Vitek, J.D.; Dangol, V. Mapping Landslide Hazards in Western Nepal: Comparing Qualitative and Quantitative Approaches. Environ. Eng. Geosci. 2010, 16, 127–142. [Google Scholar] [CrossRef]
  14. McColl, S.T. Landslide Causes and Triggers. In Landslide Hazards, Risks, and Disasters; Elsevier: Amsterdam, The Netherlands, 2022; pp. 13–41. [Google Scholar]
  15. Meena, S.R.; Puliero, S.; Bhuyan, K.; Floris, M.; Catani, F. Assessing the Importance of Conditioning Factor Selection in Landslide Susceptibility for the Province of Belluno (Region of Veneto, Northeastern Italy). Nat. Hazards Earth Syst. Sci. 2022, 22, 1395–1417. [Google Scholar] [CrossRef]
  16. Shano, L.; Raghuvanshi, T.K.; Meten, M. Landslide Susceptibility Evaluation and Hazard Zonation Techniques–A Review. Geoenviron. Disasters 2020, 7, 18. [Google Scholar]
  17. Das, J.; Saha, P.; Mitra, R.; Alam, A.; Kamruzzaman, M. GIS-Based Data-Driven Bivariate Statistical Models for Landslide Susceptibility Prediction in Upper Tista Basin, India. Heliyon 2023, 9, e16186. [Google Scholar] [PubMed]
  18. Schicker, R.; Moon, V. Comparison of Bivariate and Multivariate Statistical Approaches in Landslide Susceptibility Mapping at a Regional Scale. Geomorphology 2012, 161, 40–57. [Google Scholar]
  19. Yalcin, A. A Geotechnical Study on the Landslides in the Trabzon Province, NE, Turkey. Appl. Clay Sci. 2011, 52, 11–19. [Google Scholar] [CrossRef]
  20. Li, Y.; Liu, X.; Han, Z.; Dou, J. Spatial Proximity-Based Geographically Weighted Regression Model for Landslide Susceptibility Assessment: A Case Study of Qingchuan Area, China. Appl. Sci. 2020, 10, 1107. [Google Scholar] [CrossRef]
  21. Mandal, K.; Saha, S.; Mandal, S. Applying Deep Learning and Benchmark Machine Learning Algorithms for Landslide Susceptibility Modelling in Rorachu River Basin of Sikkim Himalaya, India. Geosci. Front. 2021, 12, 101203. [Google Scholar]
  22. Lee, S.; Talib, J.A. Probabilistic Landslide Susceptibility and Factor Effect Analysis. Environ. Geol. 2005, 47, 982–990. [Google Scholar] [CrossRef]
  23. Ghimire, M. Landslide Occurrence and Its Relation with Terrain Factors in the Siwalik Hills, Nepal: Case Study of Susceptibility Assessment in Three Basins. Nat. Hazards 2011, 56, 299–320. [Google Scholar] [CrossRef]
  24. Pourghasemi, H.R.; Teimoori Yansari, Z.; Panagos, P.; Pradhan, B. Analysis and Evaluation of Landslide Susceptibility: A Review on Articles Published during 2005–2016 (Periods of 2005–2012 and 2013–2016). Arab. J. Geosci. 2018, 11, 193. [Google Scholar] [CrossRef]
  25. Gautam, P.; Kubota, T.; Aditian, A. Evaluating Underlying Causative Factors for Earthquake-Induced Landslides and Landslide Susceptibility Mapping in Upper Indrawati Watershed, Nepal. Geoenviron. Disasters 2021, 8, 30. [Google Scholar] [CrossRef]
  26. Pradhan, B.; Oh, H.J.; Buchroithner, M. Weights-of-Evidence Model Applied to Landslide Susceptibility Mapping in a Tropical Hilly Area. Geomat. Nat. Hazards Risk 2010, 1, 199–223. [Google Scholar] [CrossRef]
  27. Kavzoglu, T.; Kutlug Sahin, E.; Colkesen, I. Selecting Optimal Conditioning Factors in Shallow Translational Landslide Susceptibility Mapping Using Genetic Algorithm. Eng. Geol. 2015, 192, 101–112. [Google Scholar] [CrossRef]
  28. Sharma, K.; Saraf, A.K.; Das, J.; Baral, S.S.; Borgohain, S.; Singh, G. Mapping and Change Detection Study of Nepal-2015 Earthquake Induced Landslides. J. Indian Soc. Remote Sens. 2018, 46, 605–615. [Google Scholar] [CrossRef]
  29. Regmi, A.D.; Devkota, K.C.; Yoshida, K.; Pradhan, B.; Pourghasemi, H.R.; Kumamoto, T.; Akgun, A. Application of Frequency Ratio, Statistical Index, and Weights-of-Evidence Models and Their Comparison in Landslide Susceptibility Mapping in Central Nepal Himalaya. Arab. J. Geosci. 2014, 7, 725–742. [Google Scholar] [CrossRef]
  30. Pandey, V.K.; Pourghasemi, H.R.; Sharma, M.C. Landslide Susceptibility Mapping Using Maximum Entropy and Support Vector Machine Models along the Highway Corridor, Garhwal Himalaya. Geocarto Int. 2020, 35, 168–187. [Google Scholar] [CrossRef]
  31. Hearn, G.J.; Hart, A.B. Landslide Susceptibility Mapping: A Practitioner’s View. Bull. Eng. Geol. Environ. 2019, 78, 5811–5826. [Google Scholar]
  32. Jebur, M.N.; Pradhan, B.; Tehrany, M.S. Optimization of Landslide Conditioning Factors Using Very High-Resolution Airborne Laser Scanning (LiDAR) Data at Catchment Scale. Remote Sens. Environ. 2014, 152, 150–165. [Google Scholar] [CrossRef]
  33. Nhu, V.-H.; Shirzadi, A.; Shahabi, H.; Singh, S.K.; Al-Ansari, N.; Clague, J.J.; Jaafari, A.; Chen, W.; Miraki, S.; Dou, J. Shallow Landslide Susceptibility Mapping: A Comparison between Logistic Model Tree, Logistic Regression, Naïve Bayes Tree, Artificial Neural Network, and Support Vector Machine Algorithms. Int. J. Environ. Res. Public. Health 2020, 17, 2749. [Google Scholar] [CrossRef]
  34. Palazzolo, N.; Peres, D.J.; Bordoni, M.; Meisina, C.; Creaco, E.; Cancelliere, A. Improving Spatial Landslide Prediction with 3d Slope Stability Analysis and Genetic Algorithm Optimization: Application to the Oltrepò Pavese. Water 2021, 13, 801. [Google Scholar] [CrossRef]
  35. Tian, Y.; Xiao, C.; Liu, Y.; Wu, L. Effects of Raster Resolution on Landslide Susceptibility Mapping: A Case Study of Shenzhen. Sci. China Ser. E Technol. Sci. 2008, 51, 188–198. [Google Scholar] [CrossRef]
  36. Zheng, B.; Wang, J.; Feng, T.; Wang, W.; Wang, Y.; Yang, Y.; Wang, G. Improving the Landslide Susceptibility Prediction Accuracy by Using Genetic Algorithm Optimized Machine Learning Approach. Int. J. Intell. Syst. 2023, 2023, 5525793. [Google Scholar]
  37. Arnone, E.; Francipane, A.; Scarbaci, A.; Puglisi, C.; Noto, L.V. Effect of Raster Resolution and Polygon-Conversion Algorithm on Landslide Susceptibility Mapping. Environ. Model. Softw. 2016, 84, 467–481. [Google Scholar] [CrossRef]
  38. Catani, F.; Lagomarsino, D.; Segoni, S.; Tofani, V. Landslide Susceptibility Estimation by Random Forests Technique: Sensitivity and Scaling Issues. Nat. Hazards Earth Syst. Sci. 2013, 13, 2815–2831. [Google Scholar]
  39. Meena, S.R.; Nachappa, T.G. Impact of Spatial Resolution of Digital Elevation Model on Landslide Susceptibility Mapping: A Case Study in Kullu Valley, Himalayas. Geosciences 2019, 9, 360. [Google Scholar] [CrossRef]
  40. Phillips, S.B.; Aneja, V.P.; Kang, D.; Arya, S.P. Modelling and Analysis of the Atmospheric Nitrogen Deposition in North Carolina. Int. J. Glob. Environ. Issues 2006, 6, 231–252. [Google Scholar] [CrossRef]
  41. Merow, C.; Smith, M.J.; Silander, J.A., Jr. A Practical Guide to MaxEnt for Modeling Species’ Distributions: What It Does, and Why Inputs and Settings Matter. Ecography 2013, 36, 1058–1069. [Google Scholar]
  42. Chen, W.; Li, W.; Hou, E.; Zhao, Z.; Deng, N.; Bai, H.; Wang, D. Landslide Susceptibility Mapping Based on GIS and Information Value Model for the Chencang District of Baoji, China. Arab. J. Geosci. 2014, 7, 4499–4511. [Google Scholar] [CrossRef]
  43. Reichenbach, P.; Busca, C.; Mondini, A.C.; Rossi, M. The Influence of Land Use Change on Landslide Susceptibility Zonation: The Briga Catchment Test Site (Messina, Italy). Environ. Manag. 2014, 54, 1372–1384. [Google Scholar] [CrossRef]
  44. Javidan, N.; Kavian, A.; Pourghasemi, H.R.; Conoscenti, C.; Jafarian, Z.; Rodrigo-Comino, J. Evaluation of Multi-Hazard Map Produced Using MaxEnt Machine Learning Technique. Sci. Rep. 2021, 11, 6496. [Google Scholar] [CrossRef]
  45. Soucy La Roche, R.; Godin, L.; Cottle, J.M.; Kellett, D.A. Tectonometamorphic Evolution of the Tip of the Himalayan Metamorphic Core in the Jajarkot Klippe, West Nepal. J. Metamorph. Geol. 2019, 37, 239–269. [Google Scholar] [CrossRef]
  46. Baral, B.; Pokharel, A.; Basnet, D.R.; Magar, G.B.; Shah, K.B. A First Photographic Record of a Yellow-Bellied Weasel Mustela Kathiah Hodgson, 1835 (Mammalia: Carnivora: Mustelidae) from Western Nepal. J. Threat. Taxa 2019, 11, 14753–14756. [Google Scholar] [CrossRef]
  47. Shahi, Y.B.; Kadel, S.; Dangi, H.; Adhikari, G.; KC, D.; Paudyal, K.R. Geological Exploration, Landslide Characterization and Susceptibility Mapping at the Boundary between Two Crystalline Bodies in Jajarkot, Nepal. Geotechnics 2022, 2, 1059–1083. [Google Scholar] [CrossRef]
  48. KC, R.; Pokhrel, R.; Acharya, P.; Sharma, K.; Subedi, M.; Bhandari, S.; Lamichhane, K. Geotechnical Assessment of the 2023 Jajarkot Nepal Earthquake Using Field Observations and Remote Sensing. Bull. Earthq. Eng. 2025, 1–25. [Google Scholar] [CrossRef]
  49. Yang, X.; Liu, R.; Yang, M.; Chen, J.; Liu, T.; Yang, Y.; Chen, W.; Wang, Y. Incorporating Landslide Spatial Information and Correlated Features among Conditioning Factors for Landslide Susceptibility Mapping. Remote Sens. 2021, 13, 2166. [Google Scholar] [CrossRef]
  50. Li, Z.; Zhu, C.; Gold, C. Digital Terrain Modeling: Principles and Methodology; CRC Press: Boca Raton, FL, USA, 2004; ISBN 0429205074. [Google Scholar]
  51. Moore, I.D.; Grayson, R.B. Terrain-based Catchment Partitioning and Runoff Prediction Using Vector Elevation Data. Water Resour. Res. 1991, 27, 1177–1191. [Google Scholar] [CrossRef]
  52. Pradhan, B.; Lee, S. Landslide Susceptibility Assessment and Factor Effect Analysis: Backpropagation Artificial Neural Networks and Their Comparison with Frequency Ratio and Bivariate Logistic Regression Modelling. Environ. Model. Softw. 2010, 25, 747–759. [Google Scholar] [CrossRef]
  53. Karra, K.; Kontgis, C.; Statman-Weil, Z.; Mazzariello, J.C.; Mathis, M.; Brumby, S.P. Global Land Use/Land Cover with Sentinel 2 and Deep Learning. In Proceedings of the 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS, Brussels, Belgium, 11–16 July 2021; IEEE: Piscataway, NJ, USA, 2021; pp. 4704–4707. [Google Scholar]
  54. Tien Bui, D.; Tuan, T.A.; Klempe, H.; Pradhan, B.; Revhaug, I. Spatial Prediction Models for Shallow Landslide Hazards: A Comparative Assessment of the Efficacy of Support Vector Machines, Artificial Neural Networks, Kernel Logistic Regression, and Logistic Model Tree. Landslides 2016, 13, 361–378. [Google Scholar] [CrossRef]
  55. Sujatha, E.R.; Sridhar, V. Landslide Susceptibility Analysis: A Logistic Regression Model Case Study in Coonoor, India. Hydrology 2021, 8, 41. [Google Scholar] [CrossRef]
  56. Cama, M.; Conoscenti, C.; Lombardo, L.; Rotigliano, E. Exploring Relationships between Grid Cell Size and Accuracy for Debris-Flow Susceptibility Models: A Test in the Giampilieri Catchment (Sicily, Italy). Environ. Earth Sci. 2016, 75, 238. [Google Scholar] [CrossRef]
  57. Arabameri, A.; Saha, S.; Roy, J.; Chen, W.; Blaschke, T.; Tien Bui, D. Landslide Susceptibility Evaluation and Management Using Different Machine Learning Methods in the Gallicash River Watershed, Iran. Remote Sens. 2020, 12, 475. [Google Scholar] [CrossRef]
  58. Reiss, H.; Cunze, S.; König, K.; Neumann, H.; Kröncke, I. Species Distribution Modelling of Marine Benthos: A North Sea Case Study. Mar. Ecol. Prog. Ser. 2011, 442, 71–86. [Google Scholar]
  59. Shi, J. Engineering Application of Landslide Susceptibility Mapping in Linear Engineering Optimization. Front. Earth Sci. 2022, 10, 861057. [Google Scholar] [CrossRef]
  60. Wang, Q.; Wang, D.; Huang, Y.; Wang, Z.; Zhang, L.; Guo, Q.; Chen, W.; Chen, W.; Sang, M. Landslide Susceptibility Mapping Based on Selected Optimal Combination of Landslide Predisposing Factors in a Large Catchment. Sustainability 2015, 7, 16653–16669. [Google Scholar] [CrossRef]
  61. Du, G.; Zhang, Y.; Iqbal, J.; Yang, Z.; Yao, X. Landslide Susceptibility Mapping Using an Integrated Model of Information Value Method and Logistic Regression in the Bailongjiang Watershed, Gansu Province, China. J. Mt. Sci. 2017, 14, 249–268. [Google Scholar]
  62. Pacheco Quevedo, R.; Velastegui-Montoya, A.; Montalván-Burbano, N.; Morante-Carballo, F.; Korup, O.; Daleles Rennó, C. Land Use and Land Cover as a Conditioning Factor in Landslide Susceptibility: A Literature Review. Landslides 2023, 20, 967–982. [Google Scholar] [CrossRef]
  63. Zhao, J.; Wang, Z.; Dong, Y.; Yang, Z.; Govers, G. How Soil Erosion and Runoff Are Related to Land Use, Topography and Annual Precipitation: Insights from a Meta-Analysis of Erosion Plots in China. Sci. Total Environ. 2022, 802, 149665. [Google Scholar] [CrossRef]
  64. Pathak, L.; Joshi, K.; Ghimire, P. Estimation of Soil Erosion Using the Revised Universal Soil Loss Equation (RUSLE) in Relation to Landslides in Mid-Hills of Nepal. J. Environ. Sci. 2023, 9, 82–93. [Google Scholar] [CrossRef]
  65. Weiss, A. Topographic Position and Landforms Analysis. In Proceedings of the Poster Presentation, ESRI User Conference, San Diego, CA, USA, 9–13 July 2001; Volume 200. [Google Scholar]
  66. Li, H.; Song, K.; Zhai, X.; Liang, M. Landslide Susceptibility Mapping Using Multiple Combination Weighting Determination: A Case Study of Collector Roads in Pingshan County, Hebei Province, China. Front. Earth Sci. (Lausanne) 2024, 12, 1403830. [Google Scholar]
  67. Giordan, D.; Cignetti, M.; Baldo, M.; Godone, D. Relationship between Man-Made Environment and Slope Stability: The Case of 2014 Rainfall Events in the Terraced Landscape of the Liguria Region (Northwestern Italy). Geomat. Nat. Hazards Risk 2017, 8, 1833–1852. [Google Scholar]
  68. Chen, Z.; Ye, F.; Fu, W.; Ke, Y.; Hong, H. The Influence of DEM Spatial Resolution on Landslide Susceptibility Mapping in the Baxie River Basin, NW China. Nat. Hazards 2020, 101, 853–877. [Google Scholar]
  69. Claessens, L.; Heuvelink, G.B.M.; Schoorl, J.M.; Veldkamp, A. DEM Resolution Effects on Shallow Landslide Hazard and Soil Redistribution Modelling. Earth Surf. Process. Landf. J. Br. Geomorphol. Res. Group. 2005, 30, 461–477. [Google Scholar]
Geosciences 15 00131 g001
Figure 1. Location of the study area (Jajarkot District) draped over the elevation and landslide polygons.
Figure 1. Location of the study area (Jajarkot District) draped over the elevation and landslide polygons.
Geosciences 15 00131 g001
Geosciences 15 00131 g002
Figure 2. Existing landslide at Gaiguwa in Barekot Rural Municipality. Photo: Badri Baral.
Figure 2. Existing landslide at Gaiguwa in Barekot Rural Municipality. Photo: Badri Baral.
Geosciences 15 00131 g002
Geosciences 15 00131 g003
Figure 3. The methodological framework of the study.
Figure 3. The methodological framework of the study.
Geosciences 15 00131 g003
Geosciences 15 00131 g004
Figure 4. Maps of CFs: (a) aspect, (b) elevation, (c) profile curvature, (d) plane curvature, (e) slope, (f) TPI, (g) proximity to rivers, (h) drainage density, (i) SPI, (j) TWI, (k) proximity to roads, (l) land use, (m) proximity to thrust, (n) dominant soil (o) lithology, and (p) landslide inventory.
Figure 4. Maps of CFs: (a) aspect, (b) elevation, (c) profile curvature, (d) plane curvature, (e) slope, (f) TPI, (g) proximity to rivers, (h) drainage density, (i) SPI, (j) TWI, (k) proximity to roads, (l) land use, (m) proximity to thrust, (n) dominant soil (o) lithology, and (p) landslide inventory.
Geosciences 15 00131 g004
Geosciences 15 00131 g005
Figure 5. Estimates of the relative contributions of the CFs in susceptibility mapping using the Maxent model.
Figure 5. Estimates of the relative contributions of the CFs in susceptibility mapping using the Maxent model.
Geosciences 15 00131 g005
Geosciences 15 00131 g006
Figure 6. Landslide susceptibility maps at different resolutions (a) landslides over hill shade, (b) susceptibility map at 200 × 200 m resolution, (c) susceptibility map at 100 × 100 m resolution, (d) susceptibility map at 50 × 50 m resolution, (e) susceptibility map at 25 × 25 m resolution, and (f) susceptibility map at 12.5 × 12.5 m resolution.
Figure 6. Landslide susceptibility maps at different resolutions (a) landslides over hill shade, (b) susceptibility map at 200 × 200 m resolution, (c) susceptibility map at 100 × 100 m resolution, (d) susceptibility map at 50 × 50 m resolution, (e) susceptibility map at 25 × 25 m resolution, and (f) susceptibility map at 12.5 × 12.5 m resolution.
Geosciences 15 00131 g006
Geosciences 15 00131 g007
Figure 7. The AUCs of the landslide susceptibility map datasets at different resolutions.
Figure 7. The AUCs of the landslide susceptibility map datasets at different resolutions.
Geosciences 15 00131 g007
Table 1. VIFs with and without TRI.
Table 1. VIFs with and without TRI.
S.N.Conditioning FactorVIF (With TRI)VIF (Without TRI)
1.Aspect1.011.01
2.Elevation2.742.72
3.Profile curvature2.132.09
4.Plane curvature2.462.35
5.Slope15.721.41
6.TPI1.991.99
7.Proximity to rivers1.671.63
8.Drainage density1.731.74
9.SPI1.101.11
10.TWI2.102.16
11.TRI15.09-
12.Proximity to roads2.001.97
13.Land use1.211.19
14.Proximity to thrust1.011.02
15.Dominant soil1.541.59
16.Lithology1.161.13
Table 2. The fitness scores of the models using the GA, based on Mahalonobis distance.
Table 2. The fitness scores of the models using the GA, based on Mahalonobis distance.
Number of FactorsSelected FactorsFitness Score (%)
4Land use, SPI, Lithology, Proximity to rivers 88.54
5Slope, Land use, SPI, Lithology, Proximity to rivers88.61
6Slope, Land use, SPI, Lithology, Proximity to roads, Aspect90.62
7Slope, Land use, SPI, Lithology, Proximity to roads, Aspect, Proximity to rivers91.03
8Slope, Land use, Drainage density, Dominant soil, Lithology, Proximity to roads, Aspect, Proximity to rivers91.26
9Slope, Land use, Elevation, Drainage density, Dominant soil, Lithology, Proximity to Roads, Aspect, Profile curvature92.46
10Slope, Land use, Elevation, Drainage density, Dominant soil, Lithology, Proximity to roads, Aspect, Profile curvature, Elevation92.03
11Slope, Land use, Lithology, Proximity to roads, Aspect, Proximity to rivers, Plane curvature, Proximity to thrust, SPI, TWI, Elevation89.98
12Slope, Land use, Dominant soil, Lithology, Proximity to roads, Aspect, Proximity to rivers, Plane curvature, Proximity to thrust, SPI, TWI, Elevation92.81
13Slope, Land use, Drainage density, Lithology, Proximity to roads, Aspect, Proximity to rivers, Plane curvature, Proximity to thrust, SPI, TWI, Elevation, TPI90.12
14Slope, Land use, Drainage density, Dominant soil, Lithology, Proximity to roads, Aspect, Proximity to rivers, Plane curvature, Proximity to thrust, TPI, SPI, TWI, Elevation91.39
15Slope, Land use, Drainage density, Dominant soil, Lithology, Proximity to roads, Aspect, Proximity to rivers, Profile curvature, Plane curvature, Proximity to thrust, TPI, SPI, TWI, Elevation91.46
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Pathak, L.; Baral, B.; Joshi, K.; Basnet, D.R.; Godone, D. Landslides in the Himalayas: The Role of Conditioning Factors and Their Resolution in Susceptibility Mapping. Geosciences 2025, 15, 131. https://doi.org/10.3390/geosciences15040131

AMA Style

Pathak L, Baral B, Joshi K, Basnet DR, Godone D. Landslides in the Himalayas: The Role of Conditioning Factors and Their Resolution in Susceptibility Mapping. Geosciences. 2025; 15(4):131. https://doi.org/10.3390/geosciences15040131

Chicago/Turabian Style

Pathak, Lalit, Badri Baral, Kamana Joshi, Dipak Raj Basnet, and Danilo Godone. 2025. "Landslides in the Himalayas: The Role of Conditioning Factors and Their Resolution in Susceptibility Mapping" Geosciences 15, no. 4: 131. https://doi.org/10.3390/geosciences15040131

APA Style

Pathak, L., Baral, B., Joshi, K., Basnet, D. R., & Godone, D. (2025). Landslides in the Himalayas: The Role of Conditioning Factors and Their Resolution in Susceptibility Mapping. Geosciences, 15(4), 131. https://doi.org/10.3390/geosciences15040131

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop