Comparative Analysis of Deterministic and Semiquantitative Approaches for Shallow Landslide Risk Modeling in Rwanda

Area of Science:

• Geosciences
• Environmental Science
• Disaster Management

Background:

• Effective landslide risk mapping is crucial for disaster management.
• Rwanda faces significant landslide hazards, necessitating robust predictive models.

Purpose of the Study:

• To compare the performance of the stability index mapping (SINMAP) and spatial multicriteria evaluation (SMCE) models for landslide risk assessment in Rwanda.
• To evaluate the predictive capabilities of both models using statistical validation techniques.

Main Methods:

• Employed SINMAP using digital elevation models and soil parameters to calculate the factor of safety.
• Utilized SMCE incorporating six landslide conditioning factors.
• Validated models using 155 historical landslide locations via receiver operating characteristic (ROC) analysis and statistical estimators (accuracy, precision, RMSE).

Main Results:

• The SMCE model outperformed SINMAP, achieving a higher area under the curve (AUC) of 0.883 versus 0.798.
• SMCE yielded superior accuracy (0.770) and precision (0.734) compared to SINMAP.
• SMCE demonstrated a lower root mean square error (RMSE) of 0.332, indicating better predictive performance than SINMAP (0.398).

Conclusions:

• The spatial multicriteria evaluation (SMCE) model shows greater efficacy in landslide risk prediction for Rwanda.
• Both SINMAP and SMCE are identified as valuable tools for landslide risk assessment in central-east Africa.
• The findings support the integration of advanced modeling techniques for improved landslide disaster management planning.