My research focuses on decentralized coordination of information sharing in disaster early warning and response. Right before, during and after disasters communities tend to self-organize to address unmet needs. However, the lack of coordination and information sharing across different community-led efforts and with the professional responding organizations can lead to duplication of efforts and other inefficiencies. Making information sharing more effective among these actors is a way to improve such coordination, empower communities and increase their resilience to disasters. The objective of my research is to provide design guidelines for effective decentralized ad-hoc coordination structures and information mechanisms in disaster early warning and response. Qualitative research studies, field research, agent-based modelling and scenario analysis are used to iteratively study the requirements, design coordination structures and information mechanisms, and test their emerging behaviour in different contexts (scenarios).