Landslide susceptibility
The interactive visualisation of landslide susceptibility is designed as a tool to facilitate the analysis of areas with the highest risk of instability in the department of Caldas. Through a dynamic dashboard, visual and control components are integrated to enable researchers, institutions, and decision-makers to explore information intuitively.
A landslide susceptibility model has been implemented using a combination of machine learning and deep learning approaches, leveraging historical landslide records and multiple environmental predictors.
Landslide susceptibility is represented in a heatmap using a continuous colour scale to indicate risk levels. Low susceptibility values (0–0.33) are shown in green–blue tones, medium susceptibility values (0.34–0.66) in yellow–orange tones, and high susceptibility values (0.67–0.99) in orange–red tones.
Layers
- Susceptibility heatmap calculated using a Convolutional Neural Network (CNN)
- Historical landslide events (FNC field reports and regional datasets)