Twitter became an important channel to contribute and consume all kinds of information, especially in times of disasters, when people feel the need for fast, real-time flows of information. Given the wealth of information Twitter provides, that information can be used by practitioners and researchers alike to study what people affected by a disaster talk about, e.g., to develop a situation awareness and to coordinate disaster management accordingly. In our research, we analyze 11 million tweets that deal with hurricane Sandy, one of the strongest hurricanes that ever hit the US east coast in 2012. First, we extract the tweets by narrowing down the hurricane affected path along the US east coast, based on geo-spatial information. Further, drawing on the situation awareness literature and previous coding schemes, we analyze the nature and characteristics of the tweets. Our research reveals that there are significantly more tweets from original sources than from secondary sources and that individuals tend to share valuable personal experiences and observations at the time of disasters. In analyzing those individual level perceptions, we illustrate how one can generate situation awareness at the collective level. This situation awareness will enhance the decision-making of disaster management agencies at the time of uncertain and volatile situations.