The prediction of future locations is of enormous research interest, partly due to the fast growing number of users of pervasive devices, as well as the tons of spatiotemporal data generated by such devices. In this paper, we propose a novel enhanced Next Location prediction technique which utilizes a trajectory model called Time Mobility Context Correlation Pattern (TMC-Pattern) and sequence alignment to mine correlations between an object and its nearest neighbors. Hidden mobility patterns drawn from such correlations are utilized in the synthesis of weighted trees called TMC-Footprint trees. The weighted TMC-Footprint trees are used together with a Markov model to predict the next location of an object with an elevated accuracy of 14,3% when compared to a state-of-the-art work. Furthermore, prediction accuracy of existing next location predictors plummets rapidly if an object suddenly takes a new next location which is absent from its trajectory history. Our technique harnesses this problem using a novel notion of Surprise Path.