Recent studies have shown that micro-blogging services such as Twitter can outperform online news aggregators in reaction time to breaking events. On the other hand, disadvantages of Twitter, such as low signal-to-noise ratio, inaccuracy of geographical data, and lack of in-depth analysis prevent it from becoming a reliable news source. In this paper, we investigate how the different information sources Twitter and online news can be integrated to gain a better understanding of current incidents for emergency response management. Our initial results suggest that the combination enables both a quick event detection and a long-term tracking of the topical and geo-spatial implications of an event. To achieve situational awareness we suggest a graph-based topic analysis approach building on top of Latent Dirichlet Allocation, as well as a new glyph design for plotting categorical data on a map, which can be used to compare news topics.