Timestamped data is any data that contains a timestamp. It could range from social networking posts, e.g. tweets, and traditional documents e.g., news articles, to sale price and volume of a traded stock in financial applications e.g., market indices. Although rich in content, the volume of these data is so high that it is challenging to get insight into the bulk at hand with minimal effort. In the literature, several sophisticated studies on event detection have been made which make use of the textual as well as temporal information. However, dealing with text is computationally intensive and it requires substantial amount of time and resources. At times, we are not equipped to, and at other times it might not be necessary to expend so much of resources. In view of this, we propose a novel TempoHierarchical algorithm, which makes use of only temporal information for identifying events. In contrast to flat timeline representation adopted in most of the past studies, it generates a hierarchical structure as output which ensures progressive disclosure of information. Each successive level of the hierarchy depicts the timeline at finer granularity of time. This allows the user to drill down the segments of timeline as per his interest. Since any domain specific information is not used, the proposed algorithm is generic in nature and works well for any timestamped dataset. However we present the algorithm in the context of tweet corpus and report our findings for two datasets.