Social Media Data Analysis and Feedback for Advanced Disaster Risk Management.

Social media are more than just a one-way communication channel. Data can becollected, analyzed and contextualized to support disaster risk management.However, disaster management agencies typically use such added-valueinformation to support only their own decisions. A feedback loop betweencontextualized information and data suppliers would result in variousadvantages. First, it could facilitate the near real-time communication ofearly warnings derived from social media, linked to other sources ofinformation. Second, it could support the staff of aid organizations duringresponse operations. Based on the example of Hurricanes Harvey and Irma we showhow filtered, geolocated Tweets can be used for rapid damage assessment. Weclaim that the next generation of big data analyses will have to generateactionable information resulting from the application of advanced analyticaltechniques. These applications could include the provision of socialmedia-based training data for algorithms designed to forecast actual cycloneimpacts or new socio-economic validation metrics for seasonal climateforecasts.