Explicit Modeling and Visualization of Imperfect Information in the Context of Decision Support for Tsunami Early Warning in Indonesia

Abstract

A certainty model and appropriate visualization techniques are presented which are applied in a newly developed Decision Support System (DSS) for tsunami early warning deployed in Jakarta, Indonesia. Our decision support approach makes use of multi-sensor fusion and pre-computed tsunami scenario simulations to create situational awareness as basis for reasonable early warning. As the Indonesian coastline is prone to near-field tsunami scenarios decision making must take place under time-critical conditions based on incomplete and uncertain information. In order to reduce the probability and the consequences of a false decision, we have developed and employed a certainty model which implies a classification of imperfect information suitable for the tsunami early warning domain and the quantification of imperfect data. The model is mapped onto and supported by appropriate visual representations.