Landslide hazard mapping is one of the most essential tools for crisis management and environmental planning in watershed areas. The Lenbran Chay watershed, located on the eastern flank of the Ahar Chay River and covering an area of 8,226 hectares, represents one of its most important headwater basins. Owing to its topographic and geological characteristics, the watershed is highly susceptible to landslide occurrence. The aim of the present study is to identify the factors influencing landslide occurrence and to delineate susceptibility zones within this watershed. For this purpose, a set of natural, anthropogenic, and morphologic variables—including distance from drainage networks, lithology, soil type, slope, aspect, elevation classes, precipitation, land use, and distance from transportation routes—was selected as the conditioning factors. The CRITIC multi criteria decision making model was then employed to determine the relative weight of each factor, based on which the final landslide hazard map was produced. The results indicated that slope is the most influential factor in landslide occurrence, followed by distance from drainage networks and lithology. Model performance assessment using training and validation datasets showed that CRITIC, with coefficients of 89.13 and 83.3 respectively, provides an acceptable level of accuracy and reliability in landslide hazard zonation. Overall, the findings suggest that the CRITIC model can be effectively utilized for risk assessment and optimal land use planning in landslide prone areas.
Introduction
Landslides are major geomorphological processes that play a critical role in shaping landscape evolution in mountainous regions and have been responsible for numerous catastrophic events worldwide. Although landslides are natural phenomena that contribute to terrain formation and erosion, they remain among the most destructive natural hazards after earthquakes and floods, causing extensive damage to infrastructure, human settlements, and natural resources. In recent decades, rapid population growth and the expansion of human activities such as deforestation, overgrazing, agricultural land degradation, and unregulated construction have significantly increased both the frequency and severity of landslide occurrences.
These developments highlight the urgent need for effective natural hazard management strategies aimed at reducing local vulnerability and mitigating broader environmental and socio‑economic impacts. Such strategies must be embedded within long‑term planning frameworks through preventive measures, preparedness initiatives, and integrated risk management approaches. Addressing landslide hazards therefore requires the application of robust analytical frameworks capable of capturing the complex interactions among geomorphological, climatic, hydrological, and anthropogenic factors that govern slope instability.
In this context, the present study applies the CRITIC multi‑criteria decision‑making model to assess and map landslide hazard in the Lanbran‑Chai Watershed. This approach, which has not previously been implemented in this watershed, provides a transparent and objective mechanism for weighting and integrating diverse conditioning factors within a spatial modeling environment. By enhancing the precision of landslide susceptibility assessment, the proposed framework offers reliable, evidence‑based support for identifying high‑risk areas and guiding targeted mitigation measures and sustainable land‑use planning.
Material and Methods
 The Lenbran‑Chai sub‑Watershed is located on the eastern side of the Ahar‑Chai River and covers an area of 8226 hectares within the defined study boundary. It represents one of the main headwater branches contributing to the Ahar‑Chai River system. Geographically, the watershed lies between 46°20′07″ to 46°30′00″ E longitude and 38°28′17″ to 38°33′52″ N latitude.
To analyze landslide hazard in the watershed, available spatial and environmental datasets were first collected and examined. Based on theoretical principles, previous research, and expert knowledge, nine conditioning factors representing geological, geomorphological, climatic, environmental, and anthropogenic characteristics were selected for analysis. These factors include:
distance from drainage networks, lithology, soil type, slope gradient, slope aspect, elevation classes, precipitation, land use, and distance from roads.
Because these datasets originally had different units and measurement scales, all layers were standardized prior to integration. Continuous variables were converted into discrete classes using GIS‑based reclassification rules. Each layer was categorized into five susceptibility classes very high, high, moderate, low, and very low based on its contribution to landslide occurrence. Standardization was performed using the Reclassify tool in ArcGIS.
After preparing the layers, the CRITIC model was used to calculate the objective weight of each conditioning factor. These weights were subsequently applied to the standardized layers within ArcMap to generate the final landslide hazard map for the Lenbran‑Chay Watershed.
Results and Discussion
Standardization and integration of the conditioning factors made it possible to conduct a consistent multi‑criteria assessment of landslide susceptibility. Each layer slope, aspect, lithology, precipitation, elevation, distance to roads, distance to drainage networks, and landuse was classified into five hazard levels, ensuring comparability across all criteria.
The landslide hazard zonation map produced using the CRITIC model indicates that 5.4, 14.7, 25.8, 22.16, and 13.6 square kilometers of the watershed fall within the very high, high, moderate, low, and very low hazard classes, respectively. These spatial patterns show that the most vulnerable areas are typically located on steep slopes, in lithologically weak zones, and in areas close to drainage channels.
Weighting results reveal that slope (26%), distance from drainage networks (13%), and lithology (11%) are the most influential factors in landslide occurrence. This is consistent with the geomorphological characteristics of the region, where steep gradients and active runoff processes, combined with susceptible rock formations, significantly increase the likelihood of slope failure.
Model validation using the Area Under the ROC Curve (AUC) demonstrates strong predictive performance. The CRITIC‑based model achieved AUC values of 0.83 (training dataset) and 0.89 (validation dataset), confirming its reliability and accuracy in identifying landslide‑prone areas within the watershed.
Conclusion
The results of this study indicate that the CRITIC multi‑criteria decision‑making model is an effective tool for landslide hazard assessment in the Lenbran‑Chai Watershed. The analysis confirms that slope, distance from drainage networks, and lithology are the most influential factors controlling landslide occurrence in the region. The final hazard map shows that 6.6%, 18.04%, 31.6%, 27.09%, and 16.64% of the watershed fall within the very high, high, moderate, low, and very low hazard categories, respectively.
The high predictive accuracy of the CRITIC model, demonstrated by robust AUC values, highlights its capability for objective weighting and reliable spatial modeling. The resulting hazard zonation map can support preventive policymaking, optimized land‑use planning, and efficient resource allocation in disaster management.
Given these promising results, future research is recommended to integrate the CRITIC model with advanced analytical approaches such as machine learning techniques, ensemble methods, or hybrid decision‑making frameworks to further increase the precision and reliability of landslide susceptibility mapping.