Profiling Flood Risk through Crowdsourced Flood Level Reports

Document Type

Article

Publication Date

10-2014

Abstract

Disaster risk reduction and management, which includes flood risk management, is among the top priorities in the Philippines. In the process of contributing to flood monitoring and public awareness, FloodPatrol, an Android mobile phone application, allows the "crowd" to report flood levels in various locations. However, a crowdsourcing-based approach poses the challenge to the credibility of the crowdsourced data. Towards the goal of having a validation model using crowdsourced flood level reports, this study presents two results. First, that there is a significant difference between "No Flood" reports and "Flood-leveled" reports. Second, that there are four possible distinct groups shown in profiling the flood reports by location.

Share

COinS