The time scale is an important element of spatial-temporal data model thatdetermines the level of detail to characterize and describe the geographic information. In spatial-temporal databases, time granularity has a decisive impact on the topological spatial-temporal relationship of geographical objects. In this paper, at first a formal definition of basic elements of time has been put forward. Then based on set theory, the number axes can be divided into three parts according to the number of time of geographical event. Based on the intersection of the sets of two-time elements a formal description of the 9 intersection model of the linear time topological relationships between geographical events has been established. On this basis, the impacts of time granularity to topological relationships of linear time between geographic events also has been discussed with a case study.