Citation

Espada R. Int. J. Disaster Risk Reduct. 2019; 39: e101141.

Copyright

(Copyright © 2019, Elsevier Publishing)

DOI

10.1016/j.ijdrr.2019.101141

PMID

unavailable

Abstract

The Philippines is usually subjected by around 21 tropical cyclonic storms per year. Whilst information and communications are widely recognized and invaluable in managing natural disaster risks, integrating ICT assets in tropical cyclone risk assessment has given little or no attention. This study generally aimed to find an optimal and spatially-explicit approach of including ICT assets, together with other equally significant risk indicators, in assessing tropical cyclone risk in the Philippines using fuzzy logic, self-organizing map (SOM), thin plate spline (TPS) and Geographic Information System (GIS). Forty (40) cyclone risk indicators/variables and two (2) benchmark variables were standardized using fuzzy logic. Random points were extracted from the fuzzified datasets and then analyzed using a neural network clustering with a 2-dimension layer of 100 neurons arranged in a 10 x 10 hexagonal grid. Examination of the SOM planes revealed three decision options: accept, eliminate, and uncertain to accept or eliminate some of the cyclone risk variables. To address this uncertainty, examination of 3D surface and 2D contour plots of thin plate spline (TPS) was further performed. The accepted 22 cyclone risk indicators were then combined using the fuzzy gamma overlay operation to generate the cyclone risk map. The output map revealed that 65% of the regions in the country are classified with high to extremely high levels of risk to tropical cyclones. The results of this study can provide valuable insights for an informed policy/decision-making to improve the country's ICT infrastructures and services for reducing its vulnerability from the impacts of the aforementioned hydro-meteorological hazard.

Language: en

Keywords

Fuzzy logic; Geographic information system (GIS); Information & communications technology; Self-organizing map; Thin plate spline analysis; Tropical cyclone risk assessment