The Future of Earthquake Impact Estimation
Impact estimation requires consideration of uncertain models and data sources since the main components—namely shaking, exposure, and vulnerabilities—entail inherent uncertainties. Since actionable response or planning requires confidence in our results, improvements in our loss calculations will require continued seismological and engineering collaboration and expanded tools to reduce modeling uncertainties. Further advancements in remote sensing, rapid in-situ monitoring and impact reporting integrated with machine learning strategies will allow for data-fusion that integrates existing loss and secondary hazard models to significantly improve the accuracy and spatial resolution of rapid shaking and loss estimates. In addition, some key new contributing datasets could radically improve our loss estimate capabilities, including global building footprints and inventories, better macroseismic constraints, early reports of fatalities for Bayesian updating, and structural health monitoring. Many of the same tools and strategies needed for real-time loss estimates are also applicable for long-term loss and risk assessments.