A New Multi-Sensor Inversion Approach for Fast and Reliable Near-Field Tsunami Early Warning

The inversion of real time sensor data from diverse sources to obtain a situation perspective in short time is a hard problem. By utilizing an analog forecasting method in combination with a simultaneous evaluation of multiple sensor measurements allows for fast and accurate situation awareness. While the traditional decision matrix approach leads to frequent false warnings, the multi-sensor inversion may give more reliable results, once measurements are available. In this presentation we will describe our approach and prove its suitability by means of a benchmark experiment.