Because of the increasing number of vehicles, traffic congestion has become a significant problem for many big cities. Numeric representations of traffic congestion cause deep concern in the population, and existing indices of traffic congestion are difficult to understand. In this paper, we propose a new concept of a traffic congestion score (TCS) that is computed by using an approximation of the speed limit. To aggregate the spatiotemporal TCS, we suggest a chained computation framework that is composed of two type of Mapreduce algorithms in Hadoop.