Exploring best-fit parameters for propagation modeling of shallow landslides, using Flow-R

Approximately one-fifth of the Earth’s surface is exposed to natural hazards – including landslides – yet most studies focus on susceptibility mapping rather than propagation dynamics. Losses from landslides occur predominantly during the post-failure propagation of the failed mass, when runout accelerates, deforms internally, and often evolves into rapid debris flows. This research analyses data from the 31 March 2017 Clustered Shallow Landslide (CSL) event in Mocoa (Colombia) using the empirical Flow-R model. This model combines terrain attributes, flow-direction algorithms, and frictional laws to simulate runout based on travel-distance angle, velocity, and a spreading exponent. Calibration against validation metrics yields optimal parameters: a minimum travel-distance angle of 15°, a maximum velocity of $10{\mathrm{ms}}^{-1}$, and exponent values $x=2$ and 4. We also establish a 25% impact-probability threshold for hazard zoning. These findings advance our understanding of CSL propagation dynamics in steep tropical environments and provide practical parameter guidelines for risk management and land-use planning.