Dimensional modeling is essentially a logical modeling of business requirements. It is different from data modeling. Lets look at what's dimensional modeling, why is it different?, why is it needed to be different?, what are special situations and how do we deal with them?.
Dimensional model is equivalent of logical data design of Data Warehouse, and much more. It is more simplistic in design and suits the purpose of a data warehouse.
Dimensional modeling is different from the OLTP normalized modeling to enable analysis and querying through massive and unpredicted queries. Something which is a relational model is ill-equipped to handle.
Data Warehouse is a repository which feeds data marts, and other down stream systems. It has to be designed to have global or re-usable set of dimensions and measures.
