Tuesday, 24 January 2017: 4:15 PM
611 (Washington State Convention Center )
Severe Weather Warning systems are undergoing a paradigm shift. Several trends are driving this shift. First, is an explosion of data from all sources, such as high-resolution meteorological observations, probabilistic forecasts, cameras, road traffic information, crowd sourcing, and social media. This abundance of data is challenging because humans still have to integrate different sources of data for decision making. Second, this shift is being propelled by the proliferation of smart phones among the general population and the trend toward hyperlocal, user-selected information. Just as people now pick when and where they watch “television shows”, there is an opportunity to allow users to customize weather information to meet current context and preferences. And third, advances in networking and computation have enabled the Internet of Things, where sensors and devices share information and initiate actions and notifications.
What’s the best way to research and validate new concepts for warning systems given these trends? This presentation will discuss one approach, CityWarn. CityWarn is a next generation hazardous weather warning system that uses high spatiotemporal weather observations integrated into a Context Aware Notifications System (CNS), a state of the art networking and communications platform, to deliver geographically targeted, user defined alerts to individuals on mobile phones through a specialized app. The warning system operates during real weather events, and collects data on system performance, mobile user feedback and actions. The app maintains user locations, preferences, and can administer short mobile surveys to app holders. Thus the warning system functions both as a research infrastructure for studying warning response from sensor observation to user behavior, and a prototype innovative severe weather warning system. Recent research results and lessons learned from the operation of CityWarn in the Dallas Fort Worth Metroplex will be presented.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner