Due to the rapid development of approaches to geographic representation, dynamic field model was derived in this paper. The dynamic field model is designed by taking the temporal dimension into account. It is further classified into two sub types of DCF (Dynamic Continuous Field) and DDF (Dynamic Discrete Field), which were designed in detail as well. After bringing forward the concept model of the dynamic field, the authors implemented it in PostgreSQL and input some pollution sampling data for testing. Running effects indicate the dynamic field works well for temporal operations and enhances the geographic representation, offering simple solutions to manipulation and complex analysis to spatio-temporal objects.