High-resolution modelling of island exposure to natural hazards tested with real disasters

Abstract

Oceanic islands are multi-risk territories but statistical aggregation of socio-economic exposure data is often a constraint for high-resolution risk modelling and hazard prevention. This work presents a downscaling procedure to obtain a complete high-resolution cartographic base on the distribution of main socio-economic variables in the Canary Islands (Spain). For this purpose, a new dasymetric procedure has been developed based on the combination of cadastral censuses, detailed planimetries and LiDAR altimetry data. The methodology allowed for the construction of an exposure cartographic base (ECB) that comprises population, capital stock, productivity and heritage (cultural and natural) layers, covering the entire archipelago at 2.5 m resolution. The ECB results was tested for accuracy and found to be 90% accurate within a positional range of 50 m. The ECB was then compared with real damages in three recent natural disasters: a volcanic eruption on La Palma in 2021, a wildfire in Gran Canaria in 2019 and a coastal flooding in Tenerife in 2018. The comparison between modelled exposure and actual damage revealed the consistency of the cartographic base for full-damage events and the need to incorporate the vulnerability factor to obtain a more accurate estimate for partial damage events.