Earthquake Early Warning for Transport Lines

This paper discusses an earthquake early warning system for transport lines. The focus lies on first producing an alert map before the strong motion phase to warn endangered trains and on then providing a damage map of the railway infrastructure immediately after the strong motion phase. We find that an efficient framework for earthquake early warning is provided by combining the information from a neural network algorithm for predicting ground motion in real-time, by railway track integrated sensors, and by risk analyses for railway operation and infrastructure via a service-oriented system architecture. In particular, we demonstrate that by consistently employing standardized and end user adapted services throughout the entire early warning chain, pertinent early warning information required for efficient rescue and repair measures can be made instantly available to end users in the form of alert and damage maps.