Data warehouses (DWs) integrate data from multiple and heterogeneous data sources. Most of the DW design methods assume that the contents of the dimensions in a DW will not change, but this is not the case in reality. Therefore, DWs must reflect these changes in the real-world in order to enable users to ask various types of temporal queries. Since temporal queries are complex and costly, it is necessary to know which modeling approach is better for such queries. In this paper, we discuss two possible approaches to implement a DW capable of maintaining the history of the changes in dimension members. We also present a classification of temporal queries that can be used to evaluate the two approaches.